JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 1

More Than Data Stories: Broadening the Role of

Visualization in Contemporary Journalism

Yu Fu and John Stasko

Abstract—Data visualization and journalism are deeply connected. From early infographics to recent data-driven storytelling,

visualization has become an integrated part of contemporary journalism, primarily as a communication artifact to inform the general

public. Data journalism, harnessing the power of data visualization, has emerged as a bridge between the growing volume of data and

our society. Visualization research that centers around data storytelling has sought to understand and facilitate such journalistic

endeavors. However, a recent metamorphosis in journalism has brought broader challenges and opportunities that extend beyond

mere communication of data. We present this article to enhance our understanding of such transformations and thus broaden

visualization research’s scope and practical contribution to this evolving ﬁeld. We ﬁrst survey recent signiﬁcant shifts, emerging

challenges, and computational practices in journalism. We then summarize six roles of computing in journalism and their implications.

Based on these implications, we provide propositions for visualization research concerning each role. Ultimately, by mapping the roles

and propositions onto a proposed ecological model and contextualizing existing visualization research, we surface seven general topics

and a series of research agendas that can guide future visualization research at this intersection.

Index Terms—journalism, data visualization, computational journalism, data-driven storytelling

✦

1 INTRODUCTION

ATA interacts with journalism in many ways. One of

the most important ways is the communication of data,

frequently performed through charts and visualizations.

Newspapers were among the ﬁrst media to bring infograph-

ics to the public’s attention, with well-known examples like

USA Today’s Snapshots, which often employed embellished

and straightforward graphics [1]. More recently, enhanced

web-based interactive visualization technologies, notably

D3.js [2], have given rise to new forms to communicate

data stories. Mainstream news platforms like The Guardian,

The New York Times, and The Washington Post were early

adopters and trailblazers of such Web-based interactive and

dynamic visual communication. Following them, a new

generation of websites, such as FiveThirtyEight and the

Pudding, has emerged in the midst of this wave, progres-

sively gaining attention from broader audiences.

While data journalism exhibits promise, conventional

journalism has been facing increasing challenges and turbu-

lence [3], [4]. The story depicted in HBO’s series The News-

room about how a cable news network ﬁghts new business

models, arising citizen journalism, and online journalism to

preserve their editorial autonomy and journalistic integrity

is ﬁctitious — and some see it as an elitist presentation of

journalism — but the diverse threats faced by journalism

and its sweeping transformation are reﬂective of reality.

Over the past decades, journalism has undergone various

changes driven by multiple social and technological forces,

particularly the digitalization of the media environment.

Journalism has long been our primary source of infor-

mation, a channel that connects us to the world outside

our tangible surroundings, a social glue that binds people

together in our “imagined community” [5], [6]. In many

countries, it is closely tied to democracy and a “watchdog”

role — to scrutinize powerful institutions, be they govern-

ments or companies [7]. Journalism’s historical importance

is undeniable, as illustrated by Pulitzer’s famous remark

— “Our Republic and its press will rise or fall together” and

Burke’s reference to it as “Fourth Estate”.

However, contemporary journalism has experienced a

crisis in perceived importance and credibility, as manifested

by the decline in its overall ﬁnancial health and audience

trust [8], [9], [10], [11]. Online journalism and citizen jour-

nalism, both spawned by digital technologies, have been

assigned blame due to emphasizing speed and undermining

journalism’s professional boundaries and its conventional

pursuit of objectivity. These are not the lone challenges

— our increasingly complicated world and the explosive

growth of information transform journalists’ role into a

more challenging one — a ﬁlter, a transmitter, an organizer,

an interpreter, and a fact-deliverer [12]. Technology has also

brought opportunities — hypermedia/multimedia content

that provides better context, interactivity that allows more

agency of audiences and potentially sparks engagement,

and platforms that enable ordinary people to participate

in civic dialogues. Journalists have turned to computa-

tional technologies to counter their role transformation. The

early adoption of computer-assisted reporting leveraging

social science statistics won investigative journalists mul-

tiple Pulitzer prizes. Data journalism featuring data-driven

storytelling and interactive visualization dashboards has re-

cently garnered mainstream attention and played a vital role

in combating the pandemic and informing the public [13].

Using data to tell stories has drawn signiﬁcant atten-

tion from the visualization research community[14], [15],

[16], with inﬂuential researchers advocating for its impor-

tance [17], [18]. Data storytelling has likely become the

most prominent topic at the Journalism-Visualization inter-

section, and rightfully so. However, existing visualization

research centering around data storytelling (e.g., [16]) tends

to focus on data journalists/platforms and their audiences,

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 2

Fig. 1: An overview of the paper structure and contributions

i.e., those who are already comfortable with data communi-

cation. We consequently lack a comprehensive understand-

ing of other journalists and audiences with limited data

infrastructure, skills, experience or literacy. Furthermore,

data-driven storytelling emphasizes the communicative role

of visualization, whereas journalists’ tasks extend beyond

mere communication. In addition, the traditional perspec-

tive regarding news organizations as ﬁxed institutions does

not adequately consider today’s increasingly dynamic and

participatory information environment.

We believe it is an appropriate time to review the state of

contemporary journalism and explore how data visualiza-

tion research can contribute to the discipline’s growth and

evolution. In particular, we ask:

• In light of the recent shifts in the information land-

scape and journalistic metamorphosis, what new vi-

sualization research opportunities could emerge at

this intersection?

• How could visualization research contextualize its

work to address broader challenges in journalism?

To answer these questions, we conducted an exploratory

scoping review [19], [20] of literature from both journal-

ism study and visualization research. In doing so, we

aim to shed light upon contemporary journalistic contexts

and challenges faced by broader journalists (e.g., pub-

lic/independent journalists) and their audiences, the roles

played by computational technologies in addressing such

challenges, and the unique value that data visualization can

contribute to these roles. Our goal is to provide an updated

and holistic perspective on how visualization research could

aid journalism. This review consists of three major parts,

with each denoting a facet of our contribution, as shown

in Figure 1:

C1. In section 3, we distill a set of important changes

in journalism driven by digital technologies, along with

journalism’s computational turn in response to emerging

challenges. We summarize six roles of computational tech-

nology in contemporary journalism and explain the ratio-

nale behind and implications for computing intervention.

C2. In section 4, we review discourse on the value of

visualization and contextualize these values in journalism.

Building off these discussions on visualization values, we

offer propositions (Figure 2) concerning each of the six roles

and map them onto an ecological model (Figure 3) to further

contextualize them.

C3. In section 5, based on our propositions, we surface

seven general topics and multiple subtopics to encompass

future Journalism-Visualization interdisciplinary research.

For each subtopic, we synthesize existing visualization re-

search, contextualize the work in journalism, and identify

future research agendas.

2 METHODOLOGY

2.1 Review Methodology

Due to the complex and heterogeneous nature of the two

ﬁelds and the variety of topics to be covered, we did not

choose a systematic review or a traditional survey. Instead,

our method resembles a scoping review, a relatively new,

but increasingly popular approach for synthesizing research

evidence [19]. HCI researchers have recently adopted scop-

ing reviews on different occasions [21], [22]. Such a review

aims to “map rapidly the key concepts underpinning a research

area and the main sources and types of evidence available” and

is commonly used to understand the range and nature of

research activity, determine the value of a full systematic

review study, summarize and disseminate research ﬁndings,

and identify research gaps in the existing literature and

aid formulating future research agendas [23]. Compared to

a traditional systematic review, a scoping review tends to

address broader topics instead of seeking answers to speciﬁc

research questions. According to CIHR’s Guide to Knowledge

Synthesis, scoping reviews “often do not undertake a detailed

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 3

appraisal of identiﬁed evidence sources and detailed synthesis of

the results of the studies” [20].

We used Google Scholar as our primary search plat-

form since the literature we need to review spreads across

two ﬁelds and their sub-domains. Our review was an it-

erative process, and we progressively expanded and re-

vised our search term pool during the process. We began

by searching for literature that discusses journalism more

comprehensively, focusing on the changes and challenges

brought about by technological developments. Therefore,

we started with broader terms such as “journalism studies”,

“journalism + new media”. Since these terms often led

to vast and homogeneous literature, we tended to follow

highly cited work. In journalism, these are often books that

cover various topics. We skimmed through the books and

reviewed the chapters and content that we deemed relevant

– for example, we included Journalism, Trust, and Credibility

and excluded Journalism Education from The Handbook of

Journalism Studies [24]. We distilled themes (e.g., participa-

tory journalism, interactive journalism, news personaliza-

tion, journalistic routine, journalism and trust, etc.) from

our initial review phase, and used these themes for our

second-round search. Our third round search focused on the

computational exploration in journalism; our search terms

include area-based ones like “computer-assisted reporting”,

“precision journalism”, “data (-driven) journalism”, “com-

putational journalism”, and topic-based ones, like “auto-

mated journalism”, “algorithmic accountability + journal-

ism”, “misinformation”, “fact-checking in journalism”, and

so on. We also expanded our pool by following leading

researchers in the respective areas (e.g., Meyer, Diakopoulos,

Lewis) and using snowballing procedures [25]. This liter-

ature is mostly peer-reviewed articles. As for the InfoVis

literature, we expanded our literature pool by searching

matching visualization keywords (e.g., “augmenting video

+ visualization”) and from our personal collections. We

also searched the ACM digital library and the IEEE Xplore

library. The literature we reviewed across these two ﬁelds

and multiple topics included 187 documents. Table 1 shows

the distributions of publication venues of literature from

both ﬁelds. We submit the list as supplementary materials.

TABLE 1: Literature Venues

Publication Venue Count

IEEE TVCG 37

ACM CHI 16

ACM UIST 4

IEEE CGA 4

ACM BELIV 4

Others 28

Total 93

(a) Visualization Literature

Publication Venue Count

Digital Journalism 11

Journalism Practice 4

New Media & Society 3

Journalism 2

Journalism Studies 2

Others 72

Total 94

(b) Journalism Literature

2.2 Reﬂexivity

This study is based on existing work that spreads across

different topics in these two large ﬁelds. While we strove to

be unbiased during the process, our experiences unavoid-

ably affected how we screened, organized, and presented

the work we reviewed. The ﬁrst author has multiple years

of experience as a professional/independent journalist and

is now a visualization researcher. The second author is a

seasoned visualization researcher. Our experiences in these

two ﬁelds lend us a unique yet balanced perspective on the

challenges and opportunities faced by broader journalists

and audiences, as well as the value visualization can pro-

vide. While we present a series of potential technological

solutions, we do not endorse a technological determin-

ism/solutionism perspective. Rather, we believe visualiza-

tion technology simply offers another option that has its

unique merits and shortcomings — their impact on pub-

lic communication awaits further investigation. We would

also point out a limitation of our study — the journalism

literature we reviewed is primarily about journalism in the

Western world, but journalism in other countries may have

distinctly different roles and characteristics.

3 JOURNALISM’S DIGITAL METAMORPHOSIS AND

COMPUTATIONAL TURN

Since entering the digital age, journalism has undergone

a continuous metamorphosis. A new form of journalism

featuring ubiquitous news, global information access, in-

stantaneous reporting, interactivity, multimedia content,

and customized content has arisen [9], [26], followed by

social media, offering the promise of the “biggest audience

reach” and a “perfect public sphere where everyone can have

a voice” [27]. Such a metamorphosis pervades and affects

many aspects associated with journalism.

In this section, we ﬁrst highlight four salient transfor-

mations driven by more general developments in technol-

ogy and discuss the challenges accompanying them. Next,

drawing from digital journalism and computational journal-

ism literature, we introduce how journalism practitioners

and scholars have proactively pursued computing tech-

nologies to aid journalism and their different perspectives.

Ultimately, by synthesizing such shifts, we summarize six

roles that computational technologies play in contemporary

journalism, along with the rationales and implications.

3.1 Salient Transformations in Digital Era

Recent shifts in journalism span multiple dimensions, rang-

ing from the nature of news content to news dissemination,

from internal organizational structures to external infor-

mation environments. Speciﬁcally, we highlight these four

salient transformations:

Enriched Media Content: The transformation in news

content should not be alien to most audiences. New capa-

bilities such as interactivity, on-demand access, user con-

trol, and personalization coupled with enriched and hyper-

linked multimedia have driven digital journalistic content

to be more contextualized, navigatable, and interactive [9],

[28], [29]. Amidst such changes, one relevant instance we

highlight is interactive journalism — multilayered visual

storytelling that engages users in interaction to unfold the

narratives [30]. Such interactive features not only have the

potential to help journalists establish authority through

enhanced visuals and sophisticated data [30], but also pro-

vide enjoyment that can re-engage audiences [29], [30] and

expand ways of thinking about journalism [30].

Participatory Environment: The Internet, digital media

technologies, and social media create the conditions for

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 4

Transformations Challenges

(T1) Interactive Journalism

C1.1: creating interactive content requires different skills than those typically acquired by

traditional journalists, notably programming skills [30]

C1.2: audiences may be unaware of or less motivated to exploit rich interactivity [29]

(T2) Participatory Journal-

ism and Communication

Environment

C2.1: it weakens journalism’s occupation boundary and roles [31], upsets its previous institu-

tional power on gatekeeping [4], [32], [33], and diminishes its traditional authoritative voice [30]

C2.2: it lowers the threshold for producing and distributing news and potentially enables an

inﬂux of misinformation or even hate speech[34]

C2.3: it is challenging to enable more coherent, elevated online discussion [35]

(T3) A Loaded and Polluted

Infosphere

C3.1: it creates a polluted public sphere where information is often met with skepticism [10]

C3.2: it places a “heavy mental burden” on the audience to distinguish trustworthy information

from a sea of misinformation [10]

(T4) News Personalization

C4: inadequate awareness of such ﬁltering and prioritization could exacerbate the “ﬁlter

bubble” phenomenon or algorithmic biases [36], [37], [38]

TABLE 2: Relevant Journalistic Transformations and The Emerging Challenges

people formerly known as “the audience” to enter jour-

nalism by enabling them to gather, create, and distribute

information in a much easier and more effective manner [6],

[27], giving birth to participatory journalism, or citizen

journalism. While participatory journalism promises to en-

gage and connect people [39] and provides a more inclusive

public communication environment that encourages citizens

to play an active role [40], [41], it nevertheless weakens the

occupation boundary and roles [31], upsets journalism’s pre-

vious institutional power on gatekeeping [4], [32], [33], and

diminishes traditional journalism’s authoritative voice [30].

The public’s competence to participate in journalistic pro-

duction and civic affairs, however, varies widely [42], as do

their interests in doing so. Technological advances have also

facilitated other participatory activities, such as quizzing

(e.g., polls, questionnaires), voicing one’s own opinion, and

commenting [42]. News consumers have welcomed and

grown accustomed to a more symmetrical communication

mode [42], [43], changing the traditional journalist-audience

relationship from a one-way, asymmetrical communication

model to a two-way dialogue [42], [44].

A Loaded and Polluted Infosphere: A more complex

world and an explosive growth of data have signiﬁcantly

complicated journalists’ roles [45]. Further, the lowered

threshold has let in a ﬂood of information, laced with a

plethora of misinformation, conspiracy theories, and fake

news — people constantly need to distinguish trustworthy

information from them [45]. In addition, a more sensation-

alistic and politically biased coverage led by competitive

business and a polarized political environment has further

eroded audiences’ trust in media [9], [10]. Collectively, these

factors have molded a loaded and polluted infosphere

where information is often met with skepticism [10], making

journalism’s role as information gatekeeper more important,

yet increasingly challenging.

News Personalization: To counter the enormous vol-

ume and multiplicity of information that overwhelms audi-

ences, tech companies like Google and Meta (Facebook) are

spearheading efforts to deliver tailored news content using

personalization algorithms fed by user data [37], [46], [47].

Personalization has grown popular among news publishers

as well [48], worrying scholars as it could largely reduce

serendipitous news discovery and exposure to alternative

viewpoints [37], thus exacerbating the “echo chamber” phe-

nomenon [36], [37], [38]. While other researchers have pre-

sented counter-evidence [49], [50] or advocated the positive

impact of news personalization [48], its potential to regulate

people’s belief systems and social inﬂuence has attracted

wide attention from researchers in different ﬁelds.

Apart from the aforementioned changes, generative AI

systems (e.g., large language models) and their applications,

with their rapidly evolving ability to perform tasks once

within the purview of journalists, have started to exert a per-

vasive inﬂuence on journalism and our society as a whole.

The ensuing transformation is unfolding before our eyes

and carries great potential to revolutionize journalism anew.

On the one hand, it is foreseeable that such AI applications

(e.g., ChatGPT) will be capable of assisting journalists with

many tasks, particularly text content generation. On the

other hand, this generative AI could exacerbate information

pollution by reducing the cost and time to produce misin-

formation/disinformation.

Regardless, further systematic investigation is needed to

identify the emerging challenges and opportunities as well

as the long-term effect generative AI technologies pose to

journalism and the journalistic environment.

3.2 Understanding Journalism’s Computational Turn

Like many other ﬁelds, journalism has embarked on its own

journey of computing innovations to accommodate its mul-

tifaceted shifts. According to Coddington’s typology [51],

its computational practices have occurred in different forms:

computer-assisted reporting, data journalism, and computa-

tional journalism.

Computer-assisted Reporting (CAR): With pioneers like

Philip Meyer [12], [54], CAR marks the entry of computing

into journalism. CAR journalists transitioned from main-

frame machines to personal computers, from simply sorting

numbers or searching for information to more sophisticated

database management and statistical analysis. Particularly,

CAR enriched the content and boosted the credibility of

investigative journalism [51]. Combined with the emergence

of the Web, CAR evolved into its second generation. On

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 5

Forms Perspectives

Computer-assisted

Reporting (CAR)

- associated closely with professional investigative reporting [51], [52], [53]

- leverages scientiﬁc methods and statistics to enhance journalism’s credibility [12], [51], [54]

- harnesses the computer’s power to manage data, conduct statistical analysis [12], [51]

Data Journalism

- places data as an information source [55] and a core in the journalistic workﬂow (curation, analysis,

communication, etc.) [51], [52]

- develops news applications that allow readers to access and explore data [55], [56], [57]

- democratizes journalistic resources, tools, and methodologies [58] and invites non-journalists to

participate in public data-driven analysis and communication [51], [58]

Computational

Journalism

- leverages computer’s processing capabilities to aggregate, automate, and abstract information [59],

make sense of information, and free journalists from low-level work [60]

- critically examines computation’s inﬂuence on journalism practice, content, and reception [53], [61],

with a focus on transparency and accountability of journalism’s algorithms [62], [63]

TABLE 3: Journalism’s Computational Practices and Their Perspectives

the one hand, more general journalists began to acquire ba-

sic computer-based skills and technologies for information

gathering; on the other hand, those CAR pioneers embarked

on more sophisticated tools for data gathering and statistical

analysis. The latter form arguably evolves into the con-

temporary approaches: data journalism and computational

journalism [51], [52].

Data Journalism: The relationship between data jour-

nalism and CAR also remains debated [64]. Bounegru [58]

suggests two main differences between them. One lies in the

role of data in news production — CAR stresses using data

as a means to enhance reporting with a focus on journalistic

tasks, while data journalism emphasizes data itself and its

function within the whole journalistic workﬂow. Another

difference lies more in the overall data environment — CAR

took place when digital data was scarce, and journalists

needed to actively seek information to answer their ques-

tions, whereas now, a vast amount of data has overwhelmed

us, and making sense of it comes as a priority [58]. Data

journalism has also delved into broader journalistic prac-

tices, while CAR is more rooted in investigative journal-

ism [51]. Parasie and Dagiral suggest that data journalism

challenges traditional CAR’s epistemological model as it

attempts to decentralize the construction of moral claims

using data analysis and consider readers as “legitimate and

active contributors” [65]. It echoes Holovaty’s early calling

for journalism to abandon “the story-centric worldview” and

accept computer programming as a form of journalism [57].

Despite that scholars and practitioners championing its

importance in today’s society due to an array of beneﬁts,

including its capability to shift journalistic context, foster

the development and application of news apps, and fa-

cilitate trustworthy journalism [55], [56], data journalism

encountered its obstacles, evidenced by the shutdown of the

EveryBlock due to its inability to engage audiences — “pages

of data weren’t enough to engage the public on their own” [55].

As Kovach and Rosenstiel [38] remark, “Journalism must

make the signiﬁcant interesting and relevant” and “Engagement

should be seen as being part of journalism’s commitment to the

citizenry”, while traditional journalists rely on instinctive

judgments and superior writing, data journalism starts to

seek for solutions from information and HCI studies [66].

Computational Journalism: Computational journalism

has emerged as an academic ﬁeld to investigate how to

advance news gathering and improve journalism workﬂow

via computational technologies [67]. Hamilton and Turner

deﬁne computational journalism as “the combination of algo-

rithms, data, and knowledge from the social sciences to supplement

the accountability function of journalism” [68]. Cohen et al.[69]

prioritize computational tools that can boost journalistic

accountability and bring positive social impact. Researchers

in this ﬁeld have proposed reﬁned conceptions and focus:

Diakopoulos suggests that computational journalism focus

on the processing capabilities of computing, notably, the

capabilities to aggregate, automate, and abstract informa-

tion [59]; Flew et al. [60] underscore computational jour-

nalism’s aim to facilitate sensemaking in journalism and

its capability to free journalists from the low-level work

(i.e., discovering/obtaining facts) as well as enhance user

engagement and enable richer user interaction with news

media. Thurman [53] notes that the scope of computational

journalism has expanded over the years, with an emphasis

on human-centric considerations and examining the use of

technological artifacts and inﬂuence in journalism [70].

3.3 Roles for Computing in Journalism

Schudson lists six functions journalism has served, partic-

ularly in a democratic society: informing the public, in-

vestigating to keep power accountable, analyzing to help

citizens comprehend a complicated world, promoting so-

cial empathy to allow other viewpoints to be appreciated,

providing a public forum for dialogue among citizens, and

mobilizing people to support certain programs [71], [72],

[73]. Our previous discussions suggest that there are various

angles for computing to support such functionalities and

respond to the challenges we identiﬁed in Table 2: techno-

logical advances to make certain tasks possible, HCI per-

spectives to accommodate the speciﬁc needs of journalists,

information and communication perspectives to understand

news/information experience and how information dis-

seminates, and critical lens to examine technology’s social

impacts. Drawing on Abebe et al.’s discourse on roles for

computing in social change [74], we summarize six roles that

computing plays in journalism that go beyond the general

use of technology (e.g., smartphones, search engines) and

explain their respective emphases:

Communicator — Computer-enabled artifacts that

communicate (i.e., engage/inform/affect) insights or narra-

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 6

tives to audiences. Such artifacts underscore computing’s

contribution to enriching news content. They range from

simpler multimedia news [44] to more sophisticated inter-

active content (e.g., data-driven stories) [30]. The latter are

mostly products of interactive journalism or data journalism

and inherit their characteristics and ethos. Computing as

Communicator concerns the design of information products

for the public that can better inform and engage them.

Facilitator — Tools that facilitate news production,

including news gathering, ﬁltering, fact-checking, and par-

ticularly, content authoring. In its broader sense, Facilitator

could encompass computational applications that facilitate

various responsibilities throughout the news production

pipeline, from handwriting transcription [9], [68] to data

wrangling (e.g., Google Reﬁne) [69]. The rise of interactive

content and enriched media types is challenging the long-

standing supremacy of written text, compelling journalists

to rely more heavily on technology to create computer-

enabled content. Hence, we focus on a more speciﬁc in-

terpretation of Computing as Facilitator, which attends to

journalists’ growing demands to create enriched journalistic

products (e.g., narrative visualization) by facilitating the

authoring process. Moreover, it is important not to disre-

gard independent/citizen journalists, who are increasingly

assuming an important authoring role.

Analyzer — Tools that assist in exploring and making

sense of information, such as extracting salient information,

deriving statistical characteristics, inferring trends, and

constructing insights/narratives. While Analyzer can be

considered a specialized type of Facilitator that assists in

data analysis, we classify it as a separate category. Such tools

have existed since the early days of CAR and have been

adopted by data journalism and computational journalism

in their respective ways. Computational journalism’s per-

spective focuses on aiding journalists with analytical activi-

ties, such as combining information from various sources,

extracting information, and exploring and mining docu-

ments [68], [69]. One speciﬁc example that Cohen et al. men-

tioned is to analyze and visualize the relationships among

entities across a document collection [69]. Data journalism’s

perspective seeks to provide the public with analytical tools

to interrogate data and develop their own interpretations.

Well-known examples include data dashboards appearing

on news websites. In this context, computing’s role as an-

alyzer can overlap and be blended with its role as commu-

nicator. Computing as Analyzer cares about helping both

journalists and their audiences better analyze information

and obtain insights and how to make analytical activities

more engaging when facing the public.

Public Forum — Digital platforms and mechanisms

for dialogue among the public and serving as a common

vehicle for perspectives of different social groups. Kovach

and Rosenstiel state that “journalism has always been a forum

for public discourse” since its origin in the Greek marketplace,

but today, this role is largely ﬁlled and failed by social

media as they enable bad actors to “distort, mislead, and

overwhelm” [38]. Allowing audiences to comment has been

the primary practice for news organizations to fulﬁll this

functionality in the digital era [35], [75]. But the emergence

of participatory journalism and enriched content further

complicates this functionality. For example, how could com-

puting facilitate and elevate public discourse around a

COVID data dashboard and drive data-driven conversa-

tions? Such emerging topics continue to await investigation.

Computing as Public Forum focuses on news platform

designs that adapt well to new media forms and enable

more coherent and elevated public discussions.

Automator — Computational technology (e.g., algo-

rithms) that automates journalistic tasks (e.g., news deliv-

ery, content creation). Automator includes personalization

algorithms [37], [62], algorithms that auto-generate con-

tent/insights [76], [77], and algorithms that alert journal-

ists/audiences about emerging events/news, etc. In partic-

ular, automated journalism that transforms structured data

to publishable news stories using programs/algorithms has

gained momentum recently, due to its potential to generate

news content in scale and in a faster, less-biased, less error-

prone manner [77]. News magnates such as the Associated

Press have already been publishing automated news con-

tent [77], [78]. Furthermore, generative AI systems have

certainly fueled the development of Automator applica-

tions. Computing as Automator offers an alternative news

production paradigm by freeing journalists from tedious,

often low-level tasks. Although it simultaneously brings

complicated social and economic considerations.

Auditor — Technology that uncovers the effect of

algorithms and audits news credibility and biases. As

algorithms have inﬁltrated more facets of journalism, re-

searchers in computational journalism started to advo-

cate investigations of algorithmic accountability and trans-

parency [62], [63], [79] and using computing as diagnostic

device [61]. Another approach is to employ computing tech-

nology as a safeguard against misinformation. Lazer et al.’s

Science article identiﬁes two categories of intervention —

individual empowerment and platform-based detection and

intervention [80]. Computing as Auditor offers the potential

to detect algorithmic biases and audit single/collective news

content for credibility in a more ubiquitous manner, thus

empowering individual readers in today’s overloaded and

polluted infosphere.

4 CONTEXTUALIZING THE VALUE OF VISUALIZA-

TION IN JOURNALISM

The previous section builds up to computing technol-

ogy’s six roles in journalism and their emphases. In this

section, we explore how data visualization can serve these

roles and provide its unique set of values to journalists and

audiences from the perspectives listed in Table 3. To achieve

that, we ﬁrst explore the fundamental values ascribed to vi-

sualization by researchers/practitioners as a starting point.

4.1 Discourse on the value of visualization

Data visualization is often touted for its value in amplify-

ing human cognition, such as expanding working memory,

enhancing pattern recognition, and making information ma-

nipulable [81]. Norman suggests that visual representations

can largely impact how efﬁciently we perform tasks [82].

Fekete et al. focus on knowledge generation and propose an

economic model that deﬁnes the total cognitive beneﬁts [83].

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 7

Fig. 2: Propositions for visualization with respect to computing’s roles in journalism

Stasko remarks that the merit of visualization lies in its

capabilities to minimize the time people need to understand

the overall essence of data or discover insights about the

data, and with greater conﬁdence [84]. He also emphasizes

the importance of interaction to these values, concurring

with Elmqvist et al.’s referring to interaction as “the catalyst

for user’s dialogue with data” [85].

The visualization community has reﬂected on its roots

in computer science and its tradition of supporting informa-

tion workers with well-deﬁned analytical tasks. Traditional

visualization research often adopts a utilitarian view that

focuses on efﬁciency, insights, and comprehension [86]. The

utilitarian view sees the functional side of visualization and

worries about aesthetics undermining this functional value,

while the aesthetic view sees its artistic side and fears that

functional foci could bore or even intimidate an audience.

Over the years, visualization researchers have explored

the values of visualization using alternative lenses. For

example, Pousman et al. proposed a new subdomain termed

“Casual InfoVis”, differentiating it from the traditional view

of visualization interfaces in four aspects: user population,

usage pattern, data type, and insights [87]. They expanded

the traditional view of visualization for analytic insight, stat-

ing that casual visualizations add awareness insight, social

insight, and reﬂective insight. In other words, casual visual-

izations can provide information that is less crystalized but

“subtly useful”, information that improves understanding of

a “social group and one’s place in it”, and information that

support “self-contemplating one’s personal and idiosyncratic

thoughts” [87]. Building off Casual InfoVis, researchers have

sought to understand the broader value of visualization

in a leisure context. Danziger [88] advocated “Information

Visualization for the People”, emphasizing regular people’s

growing needs to make sense of information in everyday

life. Sprague and Tory [89] explored people’s casual encoun-

ters with visualizations and uncovered their motivations

for using visualization beyond the utilitarian view, includ-

ing entertainment, curiosity, and social activities. More re-

cently, researchers started focusing on emotional responses

or affective effects of visualizations, particularly those for

communicative purposes [86], [90], [91], [92]. Such affective

inﬂuences are tied to the aesthetic value of visualization.

In journalism, visualization is beautifully referred to as

“the functional art” by Cairo [93] as it often embodies both

functional and artistic values. Though Cairo acknowledges

that journalism has a long history of treating infographics as

“mere ornaments” that attracts audiences and underestimates

their functional value. Cohen concurs by stating that the

core value of journalistic visualization is “deeply rooted in

measurable facts” and provides that it “offers a tantalizing

opportunity for storytelling that is above all driven by facts, not

fanaticism” [94]. Data visualization has evolved well beyond

static infographics — its aesthetic values and functional

values do not have to be encapsulated within static views

simultaneously . A comprehensive understanding of how to

leverage its aesthetic value to enhance its functional value

is a key for visualization to better support journalism —

the exquisiteness of well-made visualizations can enthrall

audiences, and that is “a valuable social currency for sharing

and attracting readers” [94].

4.2 Implications of the Six Roles for Visualization

One seemingly cogent question is how do these values of

visualization relate to computing’s six roles in journalism?

And how can visualization better support or be supported

by these six roles?

First, it is helpful to distinguish journalistic content cre-

ators from recipients (we choose it over journalist/audience

classiﬁcation since journalists can be information recipients

and ‘audiences’ are able to create content in today’s par-

ticipatory environment). Journalistic content creators are

information workers and need to accomplish comparably

crisp tasks such as gathering, ﬁltering, and analyzing raw

information and then weaving insights and narratives into

a journalistic product. These tasks align better with the

functional value of visualization. Recipients, contrastingly,

do not necessarily need to engage in these tasks — their en-

counters with visualization are leisure and loosely bounded,

thus ﬁtting better with the Casual visualization notion.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 8

Fig. 3: An ecological model for mapping computing in contemporary journalism. This model consists of two primary

processes: the news production process that roughly equates to Lee et al.’s storytelling process [18] and the news

consumption process. During the production process, creators explore and analyze data using Analyzers (e.g., visual

analytic systems), organize story pieces, then produce journalistic products using Facilitators (e.g., visualization authoring

tools). The resulting products are Communicators, including infographics, narrative visualizations, or even data videos.

Automators also have the potential to complete this process by generating journalistic content directly from structured data.

The news consumption process starts when audiences encounter these products and decide to engage. Such engagement

with visualization can provide casual (i.e., awareness, social, reﬂective) insights, curiosity, or simple entertainment. Recent

data-driven storytelling that employs interactivity enables audiences to dive further and engage in preliminary exploration,

invoking visualizations’ value as Analyzers. In today’s participatory environment, if a Public Forum (participatory

opportunity) is provided, interested audiences can easily transition into the creators’ role, triggering further analytical

engagement and completing the circle.

Second, journalism is a type of public communication

— the social inﬂuence should not be overlooked, especially

in the increasingly participatory journalism environment.

Therefore, we propose to borrow Domingo et al.’s theoret-

ical model on inclusive public communication and derived

analytical grid (access/observation, selection/ﬁltering, pro-

cessing/editing, distribution, interpretation ) [27] as a lens.

Conventionally, visualizations are often employed in

journalism as Communicator — journalists use visualiza-

tions, often static infographics, to aid their traditional sto-

rytelling that “seldom lets audience explore, but explains and

conveys ideas already thought out” [95]. But the emerging

data journalism has shifted such ethos by unleashing a

certain level of exploratory freedom to audiences. Result-

ing visualization products include data dashboards, which

allow the public to explore and analyze data indepen-

dently, and data-driven stories, which integrate visualiza-

tion’s exploratory and explanatory power [96] to “lead

readers towards a valid interpretation of underlying data” [16].

The former is a public-facing Analyzer, and the latter is

a hybrid of Analyzer and Communicator. Journalist-facing

visualization Analyzer also deals with unstructured and

textual data. For example, scholars and practitioners [53],

[69] have highlighted visual analytic tools like Jigsaw [97]

and Overview [98] that can mine insights from documents.

Facilitator in this context aims to facilitate the creation

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 9

of visualization artifacts. Mainstream generic tools like

ManyEyes [99], Tableau [100], and Datawrapper have been

underlined by scholars from both ﬁelds [68], [69], [101],

[102]. However, both parties acknowledge the demands

for more appropriate tools that support and accommodate

journalism’s workﬂow [69], [102].

The linkages between the other three roles (i.e., Public

Forum, Automator, Auditor) and visualization have mostly

remained modest in practice and experimental in research.

While these three roles are not traditionally associated with

data visualization, their growing importance drives us to ex-

plore and investigate their potential marriage, beneﬁts, and

consequences. To strengthen the existing roles (i.e., Com-

municator, Analyzer, Facilitator) and explore new roles (i.e.,

Public Forum, Automator, Auditor), we offer our proposi-

tions with respect to each role (Figure 2) and map them onto

an ecological model (Figure 3) that takes into account the

distinction of news creators and recipients, visualization’s

analytical and casual values, and Domingo et al.’s public

communication model [27].

5 RESEARCH TOPICS AND AGENDAS

In this section, we present and discuss seven research

topics and their subtopics at the Journalism-Visualization

intersection. These topics are distilled from the propositions

and the ecological model we proposed. For existing roles,

we contextualize recent visualization research and explain

how it can contribute. For unexplored roles, we point out the

research gaps and speculate on potential research directions.

By doing so, we intend to provide a map for future visual-

ization research to position itself and inspire investigations

beyond current mainstream practices.

5.1 Facilitating Data Communication

Subtopic 1.1: Facilitate Narrative Visualization Author-

ing (Facilitator)

Agenda: Support sophisticated and expressive narrative

visualization creation while lowering the barrier

Literature: [103], [104], [105], [106], [107], [108], [109], [110],

[111], [112]

Supporting more sophisticated and expressive interac-

tive visualization creation while lowering the barrier for

data visualization authoring has always been a central focus

for visualization researchers. Visualization researchers have

explored different forms of tools to author engaging nar-

rative visualizations. For example, SketchStory [104] allows

users to sketch narrative visualization in a freeform way

with pen and touch interactions. Ellipsis[113] decouples nar-

rative structure from visualization creation and allows users

to create visualization scenes, add annotations and organize

storylines. Recent tools like Lyra [105], Data Illustrator [106],

and Charticulator [107], focus on expressive authoring and

distinguish themselves from prior template editors and shelf

constructions (e.g., Tableau). According to Satyanarayan et

al. [114], such authoring tools, including subsequent Lyra

2 [108] and Data Animator [109], often assume authors’

familiarity with computational thinking and are already

acquainted with datasets [114]. Journalists’ interests tend to

align with the general direction of research on visualization

authoring tools — to lower the threshold (e.g., reducing

programming, data wrangling) and raise the ceiling (e.g.,

improving expressiveness) [114]. Idyll Studio [110], based

on the Idyll markup language [115], exhibits great potential

by lending authors the freedom to write texts and binding

texts with data and variables. Its WYSIWYG-style interface

resembles journalists’ writing environment. It also supports

Vega-Lite [116] charts and dynamic layouts and reactive

triggers that can support techniques like scrollytelling. Fur-

thermore, CrossData [111] automates the establishment of

connections between textual content and underlying data,

streamlining the process of authoring data articles. Data-

Particles [112] also leverages such latent connections to fa-

cilitate authors to create animated unit visualization, further

extending the expressive capabilities of such visualization

authoring tools.

Subtopic 1.2: Bridge the gap between exploration and

presentation (Facilitator)

Agenda: Allow journalists to visually explore data, obtain

insights, then organize the insights into engaging visual

presentations (e.g., narrative visualization) in a more efﬁ-

cient and seamless way

Literature: [117], [118], [119]

A salient feature of journalists’ information intermediary

role is that they need to perform both analytical and commu-

nicative tasks. Journalists are usually knowledgeable of their

domains but less data-savvy. Therefore, data acquisition and

wrangling can prevent them from adopting information

tools — traditional journalists are more likely to interact

with information tools through direct manipulation [120]

to ﬁnd insights and construct narratives [117]. Narrative

visualization authoring tools like Ellipsis [113] and Idyll

Studio [110] do not take data analysis into consideration.

Recently, visualization researchers have advocated a more

seamless and integrated authoring pipeline [18], [117], [118],

[119] where users can perform exploratory data analysis to

discover insights and narratives, then organize them into

narrative visualizations by manipulating layouts, annotat-

ing, highlighting, or even re-conﬁguring visual elements.

Such integration presents new challenges and opportunities.

Gratzl et al. propose the CLUE model, aiming to integrate

data exploration and presentation by capturing provenance

data during exploration and then apply such provenance

data to presentation [118]. However, visual encodings opti-

mized for exploratory purposes are not necessarily optimal

for communication. Chen et al. propose a different approach

by inserting a story synthesis phase between analytics and

storytelling. It enables users to arrange and aggregate story

slices and add annotations and visual linkings [119]. In

journalism, the data exploration and insights ﬁnding can

be messy — narratives may come from different views and

datasets across different visual analytic systems, and their

story authoring often demands originality and creativity.

It is worth exploring how to support such demands while

respecting the underlying data and relationship.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 10

Subtopic 1.3: Enable, facilitate, and understand data-

driven commenting (Public Forum)

Agenda: Enable audiences of data story/dashboard to

comment with data-driven insights (i.e., their ﬁnd-

ings/interpretations from interacting with dashboards),

design interactive mechanisms to facilitate such data-

driven dialogues among commenters, and study its in-

ﬂuence on data-driven public communication

Literature: [14], [99], [117], [121], [122], [123]

While data-driven storytelling combines data explo-

ration with data communication through interactivity and

is equipped with novel techniques to keep audiences en-

gaged, we have yet to see platforms that facilitate audiences

participating in data-driven dialogues. By data-driven di-

alogues, we mean activities in which audiences comment

and interact with others using their own analytic insights.

Commenting is among the most used feature to foster user

input in online journalism, especially when it comes to

controversial topics such as politics [123]. Surveys [122]

show people comment on the news to express emotion or

opinions, add information or correct misinformation, par-

ticipate in debates, etc. In a traditional textual environment,

it is a relatively balanced communication for audiences to

respond with texts. When arguments are accompanied by

visual support, researchers have observed that comments

tend to be more analogous and supportive [117]. Although

it could be a sign of improved credibility, it can also

be interpreted as suppression of counterarguments. Data

visualization’s effect on audiences’ commenting behavior

remains understudied. How can we facilitate better data-

driven conversational experiences for audiences? The Mar-

tini Glass structure suggested by Segel and Heer [14] can

potentially prompt readers to explore, but readers still lack

an interactive mechanism to facilitate effective exchanges

of the analytic insights they discover. Viegas et al. [99]

discusses ManyEyes’ social features — it not only allowed

users to comment underneath visualizations and datasets

but also enabled them to “snapshot” a visualization’s state

or include graphical annotations into their comments. Hull-

man et al.’s study on visualization blog comments discusses

the possibility of more sophisticated commenting interfaces

that support functionalities such as comment-presentation

linking [121]. We envision further research in this direction,

including exploring the design space for data commenting,

developing prototypes to enable data commenting activities

(e.g., linking “states” of a COVID dashboard to readers’

comments and allowing others to access quickly), and ﬁeld

deployment to investigate their impact on public data dis-

cussions using such prototypes.

5.2 Understand Visualization Dashboards on News

Websites

While the previous subsection primarily focuses on the

technical perspective, this subsection places emphasis on

empirical knowledge concerning current practices and de-

sign knowledge aimed at improving existing technology,

speciﬁcally regarding visualization dashboards.

Despite the widespread presence of visualization dash-

boards across different industries, it is only recently that

they have begun to enter our daily lives. An unprece-

dented amount of visualization dashboards have been pro-

duced to combat the COVID-19 pandemic [124]. They have

played a vital role in assisting both top-down policymakers’

decision-making and informing the public about the situa-

tion from local to global levels, thus altering our awareness

and behaviors [13]. The importance of well-designed data

dashboards cannot be understated in this misinformation-

ﬂooded era. News organizations, represented by the New

York Times, have contributed a signiﬁcant portion of these

dashboards. Such visualization dashboards as news appli-

cations have become a predominant form of data journal-

ism [30], [55], [56].

Subtopic 2.1: Investigate how visualization dashboards

are created and used as a form of journalism (Analyzer,

Communicator)

Agenda: Empirically investigate how visualization dash-

boards are designed and implemented by news organiza-

tions and used by news readers and other stakeholders

(e.g., participatory journalists)

Literature: [13], [125], [126], [127]

Visualization researchers have recently studied how

dashboards are designed and used in the wild. Through

their case survey and literature review, Sarikaya et al. [125]

characterize and organize 15 factors into categories, in-

cluding purpose, audience, visual & interactive features,

and data semantics. They also advocate for visualization

researchers to engage with users in the wild and sys-

tematically study dashboard design and implementation.

Zhang et al. answer their calling by interviewing people

involved in the COVID-19 dashboard design & creation

process [13], including those from news organizations. Their

study highlights the entangled relationships among ac-

tors and suggests a sociotechnical perspective for future

investigation. Bach et al.’s recent design workshop offers

insights on dashboard design patterns and tradeoffs [127].

Such empirical studies can deepen our understanding of

design knowledge, stakeholders’ practical pain points, the

relationships among actors and technologies, and the role

that visualization dashboards play in a larger social context.

An operationalizable contribution is design guidelines

for public-facing visualization dashboards. Design guide-

lines for other users (e.g., analysts, domain experts) are

often crisp and tailored for a set of design requirements.

Design practices often require designers to “know your au-

diences”. When it comes to the public, audiences are so

diverse that designs are often geared toward the majority,

marginalizing other groups. How could interactive visu-

alization dashboards beneﬁt more groups? We envision

more ﬂexible, customizable, and adaptable design guide-

lines as Sarikaya et al. advocate [125]. Also, evaluating

visualization artifacts has already been a big challenge

in the visualization community [128], let alone evaluating

tools that are public-facing. In-the-wild, long-term evalu-

ation methods like the multi-dimensional in-depth, long-term

case studies (MILC) method [129] proposed by Shneiderman

and Plaisant could offer inspiration for future research to

develop public-facing evaluation methods.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 11

Subtopic 2.2: Explore the design space of engaging casual

audiences into analytical tasks (Analyzer)

Agenda: Use design studies as knowledge production

devices to understand how different visual/interaction

designs impact different audiences’ transition from casual

engagements into more analytical ones

Literature: [130], [131], [132], [133]

Visualization dashboards are at their best when the

power of interactivity can be fully wielded by their users.

But as noted by multiple studies [126], [134], the usage of

interactivity in online news is an “uncomfortable myth” [134].

Then how can visualization and interaction design better

engage and encourage audiences to interact and conduct

analysis? A suitable approach to investigate such topics

is through design study. In visualization research, a design

study investigates a real-world problem and its context,

design and develops a visualization system solution, vali-

dates design choices, and reﬂect on lessons learned [130],

[131]. The resulting artifacts are usually visualization pro-

totypes and design guidelines, which can be valuable for

data journalists to borrow and follow. There is a disparity,

however — data journalists desire to inform the public who

have greater variance in data literacy, prior knowledge,

and intention, while visualization design researchers have

speciﬁc domain experts in mind and would prefer crisp

domain tasks [135]. Meyer and Dykes’s renewed discus-

sion on design study draws on Research through Design

(RtD) [132], [133] and emphasizes reﬂectively generating

knowledge about “the complex, messy, nuanced, and evolving

relationships of people with data and technology” through visu-

alization design [131]. For design studies to beneﬁt public-

facing data journalism, visualization design researchers can

explore visual/interaction design space with our ecologi-

cal model (Figure 3) in mind, which means considering

recipient’s’ encounters with dashboards casual and incen-

tivize their analytical engagement and possible transition

into creators’ role. Such design studies should shy away

from a full “utilitarian perspective” that aims at “the most

effective” solutions and “in-lab” user studies that often omit

the connections among different stakeholders. Instead, de-

sign researchers can intentionally produce insights through

different visual/interaction designs and gather long-term

user feedback from different social groups.

5.3 Integrate Interactive Visualization with New Media

Subtopic 3.1: Facilitate data video authoring and aug-

menting (Facilitator)

Agenda: Facilitate data video authoring and presentation

and augment video with visualization and novel interac-

tions

Literature: [14], [136], [137], [138], [139], [140]

The importance of video content in journalism can-

not be overstated. The emergence of short-form video

platforms (e.g., TikTok) has taken its importance to an

unprecedented level [140]. Segel and Heer [14] include

Film/Video/Animation as a basic narrative visualization

genre. Nevertheless, integrating data visualization into

videos requires tremendous effort and skills, along with

a plethora of tools [136]. Visualization researchers have

made attempts to address these challenges. One route is to

develop tools that directly facilitate data video authoring.

For example, Amini et al. developed DataClips [136] to

consolidate these creation tasks into one tool, signiﬁcantly

reducing the production efforts and time while maintaining

similar quality similar to professional editing tools. Recently,

Shin et al. developed Roslingiﬁer [138] to semi-automate

the data presentation. Another route is to augment existing

video footage with visualization. It is commonly seen in

content-focused journalism domains, such as sports. Sports

journalists and analysts frequently need to employ analytics

to tell a narrative [117]. Besides the efforts from commercial

companies, Chen et al. have done extensive work on this

front. They developed VisCommentator [137] and Sporthe-

sia [139]. The former takes raw video footage and extracts

data via an ML model, allowing users to interact with ob-

jects in the video and select the data to visualize. The latter

further eases the creation process by supporting natural lan-

guage control by leveraging NLP techniques. Furthermore,

Lin et al. recently developed Omnioculars [141], which

embedded interactive visualizations into in-game video,

extending visualization’s potential to augment broadcasting

and situated in-game journalistic analysis.

Subtopic 3.2: Visualization for live streaming (Facilitator,

Communicator)

Agenda: Design expressive spectator interfaces for mobile

device interaction and study their adoption and effects

Literature: [142], [142], [143], [144], [145]

Live streaming has recently emerged as a popular form

of social interaction, and many journalists have adopted it

as a way to do journalism. Visualization in a broadcasting

setting is not a new thing — news TV channels have been

using interactive dashboards on multiple occasions (e.g.,

CNN’s Magical Wall, ESPN). Online streaming allows more

journalists to participate in this type of communication,

which Wattenberg describes as “one person will be active,

controlling the input, while others in the group will act as specta-

tors” [142]. He calls such interactive visualizations the “Ex-

pressive Spectator Interface.” Online streaming revitalizes

an old design challenge for HCI — “designing the spectator

experience” [146]. From a sole interface design perspective,

as online streaming tends to happen on the mobile end, we

speculate that it poses new research questions that align well

with recent visualization research concerning visualization

on mobile devices and multi-modal interactions for data

visualization [143], [144], [145], though audiences’ willing-

ness to spectate and journalists’ needs for such interfaces are

subject to future investigation.

5.4 Support Journalists’ Analytical Tasks through Vi-

sual Analytics

Subtopic 4.1: Support data-driven narrative construction

(Analyzer)

Agenda: Design and develop domain-driven visual an-

alytic systems that help journalists uncover novel data-

driven storylines from a growing amount of domain data

and analytics

Supporting Literature: [117], [147]

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 12

As more domains are becoming data-driven, journalists’

demand for analytics and visualization tools to uncover

novel narratives has grown in parallel, particularly within

areas that have an abundance of data and analytics, such

as ﬁnance and sports. For example, sports journalists in-

creasingly rely on analytics to construct novel and credible

narratives, and interactive visualization offers a way to

rapidly obtain analytical insights and engaging visuals for

their storytelling [117]. Zhi et al. [147] investigated how

sports writers and fans can both beneﬁt from interactive

visualization to access statistics. Fu and Stasko [117] took

consideration of the shifts triggered by sports analytics

and developed two visualization systems tailored for bas-

ketball journalists/analysts to explore data and obtain in-

sights [117]. Assisting them in such endeavors would re-

quire a deeper understanding of both journalists’ narrative

construction, workﬂow, and the characteristics of domain

data/analytics. As such, contributions of such design study

include formative investigation to identify domain jour-

nalists’ analytical tasks, visualization prototypes that lead

to rapid analysis, novel/interesting storylines, and visual

representations that can be incorporated into stories and

insights about how such interventions impact journalists’

and their audiences.

Subtopic 4.2: Support document investigation through

visual analytics (Analyzer)

Agenda: Simplify the use of such systems; reduce data

wrangling; support ﬂexible data import, integrate new

text mining techniques; support evidence marshaling

Literature: [97], [98], [148], [149], [150]

From early CAR to recent computational journalism,

investigative journalists employ various computational ap-

proaches to perform their watchdog duties. Assisting them

in making sense of quantitative data can certainly lead to

valuable storylines, as we discussed, but here we intend

to highlight another approach that is in line with tradi-

tional investigative tasks — ﬁnding leads for stories from

large numbers of textual documents. Modern-day text min-

ing/analytics technologies, notably natural language pro-

cessing (NLP), have allowed the transformation of unstruc-

tured textual data into more structured, analyzable data,

extending visualization’s capability to represent qualitative

information and help make sense of it. A cluster of visual-

ization research has been dedicated to mining insights from

a large corpus. Jigsaw [97], [148], for example, leverages

visual analytics to aid investigative analysis. Jigsaw features

ﬂexible data imports, entity identiﬁcation, and a set of inter-

connected views to help analysts identify the connections

between entities in documents in order to glean a compre-

hensive understanding of the themes and patterns across

textual documents. Brehmer et al. speciﬁcally targeted in-

vestigative journalists and closely collaborated with them.

They developed multiple versions of Overview [98], which

led to nine published investigative stories. They studied

how journalists adopt Overview and suggested simplifying

the use and reducing data wrangling. For future research,

org et al. [149] reﬂect that data ingestion that supports

“ﬂexible data import is challenging but vital” as documents to

investigate are in a great variety of sources, forms, and for-

mats. Another key is to integrate more advanced text mining

techniques that can improve entity identiﬁcation [149]. Evi-

dence marshaling, a capability to support analysts in draw-

ing connections and adding annotations, was also proposed

to be included as an important feature [149], [150].

5.5 Visualization for Journalistic Translucence

In their prescient paper on designing digital systems that

support social processes, Erickson and Kellogg propose

three building blocks of social interaction: visibility, aware-

ness, and accountability [151]. They employ visualization

to address “blindness” in their demonstration application

“Babble”. Many online communities or conversations now

have built-in socially translucent structures or functional-

ities. Our digital blindness has been reduced to a certain

degree when it comes to awareness of other people’s pres-

ence and status. When it comes to our information ingestion,

however, our vision has seemingly become narrowed — in

today’s algorithm-pervasive information spaces, our vision

often zooms in on what is on our screens and tends to

lose track of information beyond that, which can impose

walls between algorithmically classiﬁed groups and lead to

further social fragmentation. As demonstrated by “Babble”,

visualization representations have great value in reiﬁcation.

In other words, they make abstract information perceivable,

interpretable, and even actionable. Inspired by their work,

we propose to appropriate the concept of social translucence

in journalistic contexts — journalistic translucence that con-

sists of three aspects: making news personalization “visi-

ble”, raising awareness of journalistic content consump-

tion, and holding media content production accountable.

Achieving these goals, however, requires more sophisticated

computational support. Fortunately, the abundant data and

information extraction technologies (e.g., NLP, ML), along

with advances in information visualization, particularly in-

teractivity, provide the repertoire to make it possible.

Subtopic 5.1: Make news personalization “visible” (Au-

ditor)

Agenda: Leverage interactive visualization/visual ana-

lytics to interpret news personalization algorithms and

reverse engineer them to inform the public

Literature: [62], [152], [153], [154], [155], [156], [157], [158]

HCI researchers have explored the use of visualization to

raise news personalization awareness. For example, Eslami

et al. developed FeedVis [152], which employs straightfor-

ward views to demonstrate the proportion of news that

is shown or hidden by the algorithm and allows users to

adjust the algorithms based on authorship and story con-

tent. Their study also points out four paths to news aware-

ness. Additionally, leveraging visualization to enhance the

interpretability of AI and machine learning models has

been a recent trend and promising research area [153],

[154], [155], [156]. Inspired by this cluster of research, we

envision that visualization research can contribute in two

ways: within news organizations, where news personal-

ization algorithms are available to be directly audited, vi-

sualization can be applied to interpret and explain such

algorithms to decision-makers; for news audiences who do

not have access to such algorithms, visualization can help

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 13

with “reverse-engineering” algorithms and informing the

public [62], [157].

Subtopic 5.2: Raise news consumption awareness (Audi-

tor)

Agenda: Combine personal data tracking, text analytics,

and visualization to raise individual readers’ awareness

of news consumption; potentially employ personal data

art to motivate them to explore and share with others

Literature: [159], [160]

People’s personal data has been used by media platforms

for personalization, but users themselves have very limited

access to their own data, let alone the tools to support

them in making sense of it. Recent trends, nevertheless, are

revealing people’s desire to better understand themselves

using the collection of personal data. Tech companies have

attempted to visualize users’ personal data for marketing

purposes. For instance, Wrapped, Spotify’s year-end visual

review of users’ music tastes, went viral and won Webby

People’s Voice Award for Best Data Visualization [161]. The

New York Times recently launched a similar endeavor —

Story Portrait, which generates a personalized and shareable

visual portrait of the headlines of the stories from the past

two years, with colors representing different sections [162].

The capabilities to track and collect personal news intake

can provide opportunities for readers to be aware of the

news they consume. Researchers in other areas have ex-

plored providing feedback to news readers about their

news consumption. For example, Munson et al. recently

developed Balancer [163], a Chrome Extension featuring

icons with minimum visualization. Their deployment study

suggests that showing aggregated feedback can nudge some

readers toward more balanced news consumption. As we

discussed, visualization researchers have created tools to

assist corpora analysis that could provide news readers with

more sophisticated insights beyond mere political stances.

We envision that to graft these techniques and tools onto a

journalistic setting requires further examination of how to

incite audiences’ curiosity and prompt them to explore, as

audiences do not have the motivation to expose themselves

to information outside their ideological niche. Considering

people’s favor for the aestheticization of personal data [159],

[160], data art could potentially engage audiences and pro-

vide motivations for them to explore and share.

Subtopic 5.3: Hold news production accountable (Audi-

tor)

Agenda: Leverage text analytics and interactive visual-

ization and adopt action research methodology to help

activists/advocacy hold news organization accountable

Literature: [164], [165], [166], [167], [168], [169]

Similar visual text analytics tools/techniques can be

used to examine individual/institutional news production,

potentially by social activists. News practitioners and ac-

tivists have made efforts and provided inspiration. For in-

stance, the Pudding recently published a visual essay titled

When Women Make Headlines [168]. They used data scraped

from Google News and leveraged gendered language, bias

calculation, theme dictionaries, and polarity analysis to

foreground the misrepresentation of women in the news.

Similarly, VisualizeNews combined NLP and intriguing nar-

rative visualizations to examine the press coverage of the

2019 Indian general election [166]. More serious democracy

activist efforts like Hamilton 2.0 Dashboard [167] have em-

ployed interactive visualization dashboards to demonstrate

analysis of narratives and topics promoted by state-funded

media. Research in this direction can potentially adopt ac-

tion research [164], [165], [169] as a methodology.

5.6 Combat Misinformation

“Fake news” is accelerating and impairing people’s judg-

ment of information, worsening the information environ-

ment for journalism, which is already suffering from trust

crises. Combating the scourge of misinformation is essential

to our society and has attracted serious attention from vari-

ous disciplines. Additionally, large language models have

the capacity to generate more persuasive misinformation

on a massive scale [170], highlighting the urgent need for

counteracting technologies.

Visualizations have been used to present the characteris-

tics of misinformation. For example, FakeNewsTracker [171]

employs visualization interfaces to demonstrate the fake

news they detect using machine learning models. Instead of

adding to this thread, here we identify three other directions

where visualization research can perform a larger role in

combating misinformation.

Subtopic 6.1: Visualization for investigating and commu-

nicating misinformation (Analyzer, Communicator)

Agenda: Leverage visual analytics to surface the trust-

worthiness and biases of news sources/content in order

to facilitate decision-making and fact-checking endeavors

and mitigate conﬁrmation biases; employ narrative visu-

alization to communicate fact-checking results in a more

convincing and digestible way

Literature: [45], [80], [172], [173], [174], [175], [176], [177],

[178]

Fact-checking has a long history as a standard and es-

sential routine in the newsroom [45]. With the explosion of

misinformation, fact-checking has become increasingly chal-

lenging, yet critically important [80]. We identify two angles

to approach this matter with the assistance of visualization.

The ﬁrst angle concerns the analysis of fact-checking. Fact-

checkers often need to examine different attributes of the

information [45], including entities, linguistic features (e.g.,

sentiment), and metadata (e.g., sources, timestamps). Visu-

alization possesses the capability to represent such high-

dimensional information and relationships among various

entities. Coupled with interactivity and analytics, it could

enable fact-checkers to label sources, claims, and relation-

ships more efﬁciently and with fewer biases. Visualization

researchers [172], [173], [174] have contributed tools and

designs to help investigate misinformation. Future research

can investigate fact-checking workﬂow and develop an-

alytic tools that empower them to rapidly check claims

against other statements/facts or quantitative data and

assess their credibility. In addition, techniques [177], [178]

used to surface analysts’ biases and promote self-awareness

can also be integrated into such fact-checking tools.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 14

The second angle attends to the challenge of communi-

cating results. Fact-checking websites adopt various ways

to present their verdict and analysis on claims/rumors,

including visual representations (e.g., PolitiFact’s Truth-O-

Meter [179]). Despite their efforts, delivering such verdicts

can be laborious and sometimes counterproductive if done

in a threatening manner [175], [180]. There are also ini-

tiatives aiming to make fact-checking more visible [45].

We speculate narrative visualizations, with their capabil-

ities to deliver sophisticated information in an engaging

manner, could support communicating fact-checking ver-

dict/analysis in a more digestible and transparent way

(e.g., demonstrating a verdict’s relationships with mul-

tiple sources supporting/refuting it). Another promising

approach is to employ visualizations to effectively com-

municate the verdicts on data-driven/statistical claims, as

visualization can potentially provide context for the data

claims and help audiences to comprehend the verdicts. Fu-

ture research can explore the design space and evaluate its

inﬂuence on audiences’ reception of fact-checking verdicts,

regarding aspects such as engagement and persuasiveness.

Subtopic 6.2: Detect visual deception (Auditor)

Agenda: Develop end-user tools (e.g., browser extensions)

that can detect visual deception/distortion and alert news

readers with annotations and visual augmentations

Literature: [181], [182], [183], [184]

While visualizations possess communicative and even

persuasive power, they can also be used for deception [181].

There are countless notorious cases from marketing com-

panies, news organizations, and government institutions —

even the White House posted a problematic chart on Twitter

recently [185]. Regardless of their intentions, it becomes

increasingly important to alert and inform the public about

potential visual deception. Visualization design guidelines

are often about “how NOT to lie with charts” [186], [187]

but in reality, people have so many incentives not to follow

them. More researchers have begun to focus on detecting

these “graphical lies.” Pandey et al. summarize popular

distortion techniques and their effect. They classify de-

ceptive visualizations into two groups: message exaggera-

tion/understatement and message reversal [181]. Hopkins

et al. create VisualLint to alert visualization designers about

such errors. McNutt et al. introduces a concept of “visual-

ization mirages” and suggests alerting readers to potential

issues [183]. Fan et al. developed a Chrome Extension that

automatically reads and annotates potential deceptions in

line charts, making such techniques more easily accessible

for readers [182]. This opens up future research possibilities

to broaden the chart types and increase accessibility.

Subtopic 6.3: Detect text-visualization misalignment

(Auditor)

Agenda: Develop end-user tools (e.g., browser extensions)

that can point out the misalignment between visualization

and text (e.g., titles, annotations)

Literature: [188], [189], [190]

Deception can also occur when visualizations do not

match the textual descriptions or claims in titles or anno-

tations. Borkin et al.’s experiment based on eye-tracking

suggests that titles and text can greatly impact how people

recall the gist of visualization [188]. Kong et al.’s studies

demonstrate that slanted titles and misaligned annotations

can mislead audiences’ interpretation of visualizations and

have more persistent impacts on recalling than visualiza-

tions do [189], [190]. They also suggest using algorithms

and NLP methods to combat text-visualization misalign-

ment and tools to foreground such misalignment and raise

readers’ awareness [190]. We anticipate that web browser

plugins capable of detecting and notifying readers of such

misalignment would come in handy, although it presents

technical challenges to improving the accuracy, and its im-

pact on reader experience needs further examination.

5.7 Automated Visual Stories and Insights (Automator,

Communicator)

Subtopic 7.1: Automate visual data story creation (Au-

tomator)

Agenda: Enable (semi-)automated generation of more so-

phisticated and engaging data-driven storytelling and un-

derstand readers’ perception of such automated content

Literature: [76], [77], [191], [192], [193], [194], [195]

Robot-generated news has its advantages when it comes

to speed, scalability, and even objectivity, but it often

lacks sophisticated narratives and has been deemed bor-

ing and technical [77]. HCI researchers have explored au-

tomated news content beyond just text. Oh et al. devel-

oped a system NewsRobot [192] that automatically gen-

erates different types (general/personalized) and styles

(text/image/sound) of news content using sports data,

and they used generated news as a probe to understand

audiences’ perception of different combinations of auto-

mated news. Visualization researchers have studied it from

a broader perspective — auto-insights, deﬁned as “data

observations revealed by automation” by Law et al. [191].

Their review [191] found that only a few studies were

dedicated to automated data-driven storytelling. Systems

like DataShot [193], Temporal [194], and Calliope [195] are

pushing towards more sophisticated and engaging auto-

mated visual data stories while supporting human editorial

intervention. This direction still imposes great technical

challenges (e.g., extracting data insights, generating proper

visualization, linking texts to charts, supporting sophisti-

cated techniques like scrollytelling) and lacks design impli-

cations and understanding of audience perceptions.

Subtopic 7.2: Augment interactive visualization with

automated insights and guidance (Automator)

Agenda: Provide auto-generated insights and guidance

based on users’ interactions to help ordinary audiences

interpret data and visualization

Literature: [196]

With visualization dashboards permeating various areas

in journalism and reaching broader audiences, the pub-

lic’s need for assistance in interpreting these visualizations

and extracting insights also grows. Driven by commercial

companies like Automated Insights, multiple auto-insight

systems have emerged. Such systems use statistical models

to infer potentially interesting/relevant facts in the data

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 15

and generate natural language to communicate them [196].

Srinivasan et al. discuss such template-based natural lan-

guage generation systems that assist users in visualization

interpretation. They introduce Voder, a system that automat-

ically generates data facts according to visualization and

links data facts to the visualization through visual embel-

lishments (i.e., opacity, regression lines) [196]. As they note,

such tools are still in their infancy and how to contextualize

them in journalism to beneﬁt the audiences is still a target

for future investigation. Nevertheless, such systems offer the

potential to guide ordinary people in data exploration and

interpretation and ultimately increase their data literacy.

5.8 Topic Connections and Other topics

It is worth noting that not all of the topics we list are

independent. The ecological model (Figure 3) we propose

could help demonstrate their intricate relationships. Such

relationships can be synergistic — e.g., the relationship

between investigating misinformation and communicating

verdicts, two directions mentioned in Subtopic 6.1, can be

strengthened by a facilitating tool advocated in Subtopic

1.2. The relationships can also be antagonistic — e.g., au-

tomated visual storytelling technology discussed in Subtopic

7.1 could potentially weaken human-centric approaches as

it bypasses the topics associated with Analyzer (e.g., Subtopic

4.1 & 4.2) and Facilitator (e.g., Subtopic 1.1) and generates

Communicator with minimum human intervention.

In addition, our proposed set of topics is not intended

to be exhaustive. Our goal is to open up the possibilities

for this cluster of research. Other topics, such as playable

visualization [197], [198], immersive data-driven stories [199],

and archiving data-driven stories and apps [200], [201], [202],

[203], [204] are worthy topics that have been studied by

researchers and likely have applications in journalism.

6 REFLECTION

Although the main thrust of this paper is to explore the

broader role of visualization in journalism and thus provide

future agendas for visualization research to better assist con-

temporary journalism, our review of the literature on both

fronts indicates shared interests between the two research

areas and potential beneﬁts for visualization research.

As Meijer [66] notes, journalism researchers have turned

to “news experience” and considered an “expressive” angle

to study journalism. Scholars have employed new concepts

and explanatory frames, including HCI studies and infor-

mation studies, to analyze topics like people’s experience

of information and how computing-mediated interactive

experience can alter audiences’ preferences, perceptions,

and experiences of news [66]. Such goals align well with the

visualization/HCI research, though they are likely to have

different foci. We advocate closer collaborations between

these ﬁelds. It can lead to a more balanced and heteroge-

neous perspective — visualization/HCI researchers tend to

lean towards a technological or cognitive science one while

traditional journalism study is dominated by a sociological

perspective [205]. For example, both sides have shown inter-

est in data literacy of the public — visualization researchers

may have a stronger interest in how it affects people’s

interpretation of different visual representations [206], how

to better design visuals and interactions for the public

with varying levels of data literacy, and how to improve

people’s data/visualization literacy [207], while journalism

scholars may focus on how data/visualization literacy im-

pacts public discourse and its implications for journalistic

products. Visualization research can provide devices for

data gathering and cognitive bases to explain high-level

social phenomena, while journalism study offers a holistic

social context to situate the generated knowledge and make

such knowledge more useful and actionable.

In addition to collaborating with journalism scholars,

visualization research can also beneﬁt from working with

journalism practitioners. Apart from the fact that they are

top-tier stakeholders and hence the necessity to understand

their workﬂow and needs, another practical beneﬁt is that

journalists can potentially increase the visibility and appre-

ciation of visualization outside the core visualization com-

munity, which has been advocated by some visualization

researchers [208]. The value of visualization is challenging

to perceive without “seeing it” or “interacting with it”,

especially for audiences less familiar with the affordance of

interactive visualizations and/or with limited data/visual

literacy. Journalists’ platforms, whether institutional or in-

dividual, can provide vehicles for visualizations to reach

broader audiences, increase the visibility of visualization,

and potentially improve the data and visual literacy of the

masses. Journalists are natural information transmitters —

assisting journalism in a broader way offers visualization

research a chance to increase its real-world visibility — not

only putting enthralling graphics in front of people but also

demonstrating to them how they can converse with their

surrounding information in an effective yet joyful way.

7 CONCLUSION

Journalism and visualization have an intertwined relation-

ship, one that is often overshadowed by visualization’s

capability in communication. Indeed, visualization excels at

aiding data-driven storytelling, but it can also be instru-

mental to journalists and news audiences in many other

ways, with its own set of aesthetic and functional values. In

this study, we took a step back and examined journalism’s

digital metamorphosis to identify the emerging challenges

and their implications, as well as journalism’s computa-

tional exploration to address such challenges and shifts.

Based on our ﬁndings, we identiﬁed computing’s six roles in

contemporary journalism, and their corresponding focuses.

Next, we revisited discourse on the fundamental values of

visualization and proposed a set of propositions about how

visualization can support or be supported by computing’s

six roles. By situating these roles and propositions in an

ecological model, we ultimately surfaced a range of research

topics and respective agendas. Through this paper, we hope

to shed light upon the shifts in contemporary journalism

and their implications for visualization research and inspire

broader research exploration at the intersection of these two

dynamic ﬁelds.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 16

REFERENCES

[1] J. J. Otten, K. Cheng, and A. Drewnowski, “Infographics and

public policy: Using data visualization to convey complex in-

formation,” Health Affairs, vol. 34, no. 11, pp. 1901–1907, 2015.

[2] M. Bostock, V. Ogievetsky, and J. Heer, “D³ data-driven docu-

ments,” IEEE Transactions on Visualization and Computer Graphics,

vol. 17, no. 12, pp. 2301–2309, 2011.

[3] A. Bruns, “Gatekeeping, gatewatching, real-time feedback: new

challenges for journalism,” Brazilian journalism research, vol. 7,

no. 2, pp. 117–136, 2011.

[4] S. C. Lewis, K. Kaufhold, and D. L. Lasorsa, “Thinking about

citizen journalism,” Journalism Practice, vol. 4, no. 2, pp. 163–179,

2010.

[5] B. Anderson, Imagined communities: Reﬂections on the origin and

spread of nationalism. Verso books, 2006.

[6] K. Wahl-Jorgensen and T. Hanitzsch, “Journalism studies: Devel-

opments, challenges, and future directions,” in The Handbook of

Journalism Studies, 2nd ed., K. Wahl-Jorgensen and T. Hanitzsch,

Eds. Routledge, 2019, pp. 3–20.

[7] B. McNair, “Journalism and democracy,” in The Handbook of

Journalism Studies, 1st ed. Routledge, Jan 2009, pp. 257–269.

[8] C. Peters and M. Broersma, Rethinking Journalism: Trust and Par-

ticipation in a Transformed News Landscape. Routledge, 2013.

[9] J. Pavlik, Journalism and New Media. Columbia University Press,

2001.

[10] A. V. Dalen, “Journalism, trust, and credibility,” in The Handbook

of Journalism Studies, 2nd ed., K. Wahl-Jorgensen and T. Han-

itzsch, Eds. Routledge, 2019, pp. 356–371.

[11] J. M. Ladd, Why Americans Hate the Media and How It Matters.

Princeton University Press, 2012.

[12] P. Meyer, Precision Journalism: A Reporter’s Introduction to Social

Science Methods. Rowman & Littleﬁeld Publishers, 2002.

[13] Y. Zhang, Y. Sun, J. D. Gaggiano, N. Kumar, C. Andris, and A. G.

Parker, “Visualization design practices in a crisis: Behind the

scenes with covid-19 dashboard creators,” IEEE Transactions on

Visualization and Computer Graphics, vol. 29, no. 1, pp. 1037–1047,

2023.

[14] E. Segel and J. Heer, “Narrative visualization: Telling stories with

data,” IEEE Transactions on Visualization and Computer Graphics,

vol. 16, no. 6, pp. 1139–1148, 2010.

[15] J. Hullman and N. Diakopoulos, “Visualization rhetoric: Framing

effects in narrative visualization,” IEEE Transactions on Visualiza-

tion and Computer Graphics, vol. 17, no. 12, pp. 2231–2240, 2011.

[16] N. Riche, C. Hurter, N. Diakopoulos, and S. Carpendale, Data-

Driven Storytelling, ser. AK Peters Visualization Series. CRC

Press, 2018.

[17] R. Kosara and J. Mackinlay, “Storytelling: The next step for

visualization,” Computer, vol. 46, no. 5, pp. 44–50, 2013.

[18] B. Lee, N. H. Riche, P. Isenberg, and S. Carpendale, “More than

telling a story: Transforming data into visually shared stories,”

IEEE Computer Graphics and Applications, vol. 35, no. 5, pp. 84–90,

2015.

[19] M. T. Pham, A. Raji

c, J. D. Greig, J. M. Sargeant, A. Papadopou-

los, and S. A. Mcewen, “A scoping review of scoping reviews:

advancing the approach and enhancing the consistency,” Research

Synthesis Methods, vol. 5, no. 4, pp. 371–385, 2014.

[20] C. I. of Health Research, “A guide to knowledge synthesis - cihr,”

https://cihr-irsc.gc.ca/e/41382.html, Cihr-irsc.gc.ca, 2010.

[21] D. Long and B. Magerko, “What is ai literacy? competencies and

design considerations,” in Proceedings of the 2020 CHI Conference

on Human Factors in Computing Systems, ser. CHI ’20. ACM, 2020,

pp. 1–16.

[22] V. Herdel, L. J. Yamin, and J. R. Cauchard, “Above and beyond:

A scoping review of domains and applications for human-drone

interaction,” in Proceedings of the 2022 CHI Conference on Human

Factors in Computing Systems, ser. CHI ’22. ACM, 2022.

[23] H. Arksey and L. O’Malley, “Scoping studies: towards a

methodological framework,” International Journal of Social Re-

search Methodology, vol. 8, no. 1, pp. 19–32, 2005.

[24] K. Wahl-Jorgensen and T. Hanitzsch, The Handbook of Journalism

Studies, 1st ed. Routledge, 2009.

[25] C. Wohlin, “Guidelines for snowballing in systematic literature

studies and a replication in software engineering,” in Proceedings

of the 18th International Conference on Evaluation and Assessment in

Software Engineering, ser. EASE ’14. ACM, 2014.

[26] B. Scott, “A contemporary history of digital journalism,” Televi-

sion & New Media, vol. 6, no. 1, pp. 89–126, 2005.

[27] D. Domingo, T. Quandt, A. Heinonen, S. Paulussen, J. B. Singer,

and M. Vujnovic, “Participatory journalism practices in the media

and beyond,” Journalism Practice, vol. 2, no. 3, pp. 326–342, 2008.

[28] J. V

azquez-Herrero, X. L

opez-Garc

ıa, and F. Irigaray, The

Technology-Led Narrative Turn. Springer International Publishing,

2020, pp. 29–40.

[29] D. S. Chung and C. Y. Yoo, “Audience motivations for using

interactive features: Distinguishing use of different types of in-

teractivity on an online newspaper,” Mass Communication and

Society, vol. 11, no. 4, pp. 375–397, 2008.

[30] N. Usher, Interactive journalism: Hackers, data, and code. University

of Illinois Press, 2016.

[31] S. Jones and A. Bruns, Gatewatching: Collaborative Online News

Production, ser. Digital formations. P. Lang, 2005.

[32] M. Carlson and S. C. Lewis, “Boundary work,” in The Handbook of

Journalism Studies, 2nd ed., K. Wahl-Jorgensen and T. Hanitzsch,

Eds. Routledge, 2019, pp. 123–135.

[33] P. J. Shoemaker, T. P. Vos, and S. D. Reese, “Journalists as gate-

keepers,” in The Handbook of Journalism Studies, 2nd ed., K. Wahl-

Jorgensen and T. Hanitzsch, Eds. Routledge, 2019.

[34] J. Singer, D. Domingo, A. Heinonen, A. Hermida, S. Paulussen,

T. Quandt, Z. Reich, and M. Vujnovic, Participatory Journalism:

Guarding Open Gates at Online Newspapers. Wiley, 2011.

[35] R. Zamith and S. C. Lewis, “From public spaces to public sphere,”

Digital Journalism, vol. 2, no. 4, pp. 558–574, 2014.

[36] E. Powers, “My news feed is ﬁltered?” Digital Journalism, vol. 5,

no. 10, pp. 1315–1335, 2017.

[37] E. Pariser, The Filter Bubble: How the New Personalized Web Is

Changing What We Read and How We Think. Penguin Publishing

Group, 2011.

[38] B. Kovach and T. Rosenstiel, The Elements of Journalism, Revised

and Updated 4th Edition: What Newspeople Should Know and the

Public Should Expect. Crown, 2021.

[39] S. Robinson, “Traditionalists vs. convergers: Textual privilege,

boundary work, and the journalist—audience relationship in the

commenting policies of online news sites,” Convergence, vol. 16,

no. 1, pp. 125–143, 2010.

[40] A. Hermida, Mechanisms of Participation. John Wiley & Sons, Ltd,

2011, ch. 2, pp. 11–33.

[41] D. Gillmor, We the Media: Grassroots Journalism By the People, For

the People. O’Reilly Media, 2006.

[42] J. Hujanen and S. Pietik

ainen, “Interactive uses of journalism:

Crossing between technological potential and young people’s

news-using practices,” New Media & Society, vol. 6, no. 3, pp.

383–401, 2004.

[43] D. S. Chung, “Interactive Features of Online Newspapers: Iden-

tifying Patterns and Predicting Use of Engaged Readers,” Journal

of Computer-Mediated Communication, vol. 13, no. 3, pp. 658–679,

04 2008.

[44] J. Pavlik, “The impact of technology on journalism,” Journalism

Studies, vol. 1, no. 2, pp. 229–237, 2000.

[45] L. Graves, Deciding What’s True: The Rise of Political Fact-Checking

in American Journalism. Columbia University Press, 2016.

[46] A. LaFrance, “The power of personalization,” https://

niemanreports.org/articles/the-power-of-personalization/, Nie-

man Reports, Feb 2019.

[47] M. Powers, “In forms that are familiar and yet-to-be invented,”

Journal of Communication Inquiry, vol. 36, no. 1, pp. 24–43, 2012.

[48] N. Thurman and S. Schifferes, “The future of personalization at

news websites,” Journalism Studies, vol. 13, no. 5-6, pp. 775–790,

2012.

[49] M. Haim, A. Graefe, and H.-B. Brosius, “Burst of the ﬁlter

bubble?” Digital Journalism, vol. 6, no. 3, pp. 330–343, 2018.

[50] S. Flaxman, S. Goel, and J. M. Rao, “Filter Bubbles, Echo Cham-

bers, and Online News Consumption,” Public Opinion Quarterly,

vol. 80, no. S1, pp. 298–320, 03 2016.

[51] M. Coddington, “Clarifying journalism’s quantitative turn,” Dig-

ital Journalism, vol. 3, no. 3, pp. 331–348, 2015.

[52] A. Gynnild, “Journalism innovation leads to innovation jour-

nalism: The impact of computational exploration on changing

mindsets,” Journalism, vol. 15, no. 6, pp. 713–730, 2014.

[53] N. Thurman, “Computational journalism,” in The Handbook of

Journalism Studies, 2nd ed., K. Wahl-Jorgensen and T. Hanitzsch,

Eds. Routledge, 2019, pp. 180–195.

[54] P. Meyer, “Deﬁning and measuring credibility of newspapers:

Developing an index,” Journalism Quarterly, vol. 65, no. 3, pp.

567–574, 1988.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 17

[55] A. B. Howard, “The art and science of data-driven journalism,”

Academic Commons, 2014.

[56] J. Gray, L. Bounegru, and L. Chambers, The Data Journalism

Handbook: How Journalists Can Use Data to Improve the News.

O’Reilly Media, 2012.

[57] A. Holovaty, “A fundamental way newspaper sites

need to change,” http://www.holovaty.com/writing/

fundamental-change/, 2006.

[58] L. Bounegru, “Data journalism in perspective,” in The Data Jour-

nalism Handbook. O’Reilly Media, 2012.

[59] N. Diakopoulos, A Functional Roadmap for Innovation in Computa-

tional Journalism, 2011.

[60] T. Flew, C. Spurgeon, A. Daniel, and A. Swift, “The promise of

computational journalism,” Journalism Practice, vol. 6, no. 2, pp.

157–171, 2012.

[61] L. Lepp

anen, H. Tuulonen, and S. Sir

en-Heikel, “Automated

journalism as a source of and a diagnostic device for bias in

reporting,” Media and Communication, vol. 8, no. 3, pp. 39–49,

2020.

[62] N. Diakopoulos, “Algorithmic accountability,” Digital Journalism,

vol. 3, no. 3, pp. 398–415, 2015.

[63] N. Diakopoulos and M. Koliska, “Algorithmic transparency in

the news media,” Digital Journalism, vol. 5, no. 7, pp. 809–828,

2017.

[64] P. Hammond, “From computer-assisted to data-driven: Journal-

ism and big data,” Journalism, vol. 18, no. 4, pp. 408–424, 2017.

[65] S. Parasie and E. Dagiral, “Data-driven journalism and the

public good: “Computer-assisted-reporters” and “programmer-

journalists” in Chicago,” New Media & Society, vol. 15, no. 6, pp.

853–871, 2013.

[66] I. Meijer, “Journalism, audiences, and news experience,” in The

Handbook of Journalism Studies, 2nd ed., K. Wahl-Jorgensen and

T. Hanitzsch, Eds. Routledge, 2019, pp. 389–405.

[67] G. T. G. Center, “Journalism 3g: The future of technology

in the ﬁeld: A symposium on computation + journalism,”

https://web.archive.org/web/20100103125218/http://

www.computational-journalism.com/symposium/index.php,

Archive.org, 2 2008.

[68] J. T. Hamilton and F. Turner, “Accountability through algorithm:

Developing the ﬁeld of computational journalism,” in Report from

the Center for Advanced Study in the Behavioral Sciences, Summer

Workshop, 2009, pp. 27–41.

[69] S. Cohen, J. T. Hamilton, and F. Turner, “Computational journal-

ism,” Communications of the ACM, vol. 54, no. 10, pp. 66–71, 2011.

[70] S. C. Lewis and O. Westlund, “Actors, actants, audiences, and

activities in cross-media news work,” Digital Journalism, vol. 3,

no. 1, pp. 19–37, 2015.

[71] M. Schudson, Why Democracies Need an Unlovable Press. Wiley,

2008.

[72] ——, “News and democratic society: past, present, and future,”

Hedgehog Review, vol. 10, no. 2, pp. 7–21, 2008.

[73] N. Usher, “Why democracies need an unlovable press,” 2009.

[74] R. Abebe, S. Barocas, J. Kleinberg, K. Levy, M. Raghavan, and

D. G. Robinson, “Roles for computing in social change,” in

Proceedings of the 2020 Conference on Fairness, Accountability, and

Transparency, ser. FAT* ’20. ACM, 2020, pp. 252–260.

[75] J. D. Wolfgang, “Pursuing the ideal: How news website com-

menting policies structure public discourse,” Digital Journalism,

vol. 4, no. 6, pp. 764–783, 2016.

[76] M. Carlson, “The robotic reporter,” Digital Journalism, vol. 3, no. 3,

pp. 416–431, 2015.

[77] A. Graefe, “Guide to automated journalism,” Columbia Journalism

Review, 2016.

[78] L. Lepp

anen, M. Munezero, M. Granroth-Wilding, and H. Toivo-

nen, “Data-driven news generation for automated journalism,” in

Proceedings of the 10th International Conference on Natural Language

Generation. Association for Computational Linguistics, Sep. 2017,

pp. 188–197.

[79] B. Mittelstadt, “Automation, algorithms, and politics— auditing

for transparency in content personalization systems,” Interna-

tional Journal of Communication, vol. 10, no. 0, p. 12, 2016.

[80] D. M. J. Lazer, M. A. Baum, Y. Benkler, A. J. Berinsky, K. M.

Greenhill, F. Menczer, M. J. Metzger, B. Nyhan, G. Pennycook,

D. Rothschild, M. Schudson, S. A. Sloman, C. R. Sunstein, E. A.

Thorson, D. J. Watts, and J. L. Zittrain, “The science of fake news,”

Science, vol. 359, no. 6380, pp. 1094–1096, 2018.

[81] S. Card, S. Shneiderman, M. Card, J. Mackinlay, and B. Shneider-

man, Readings in Information Visualization: Using Vision to Think,

ser. Interactive Technologies. Elsevier Science, 1999.

[82] D. Norman, Things That Make Us Smart: Defending Human At-

tributes in the Age of the Machine. Diversion Books, 2014.

[83] J. Fekete, J. J. van Wijk, J. T. Stasko, and C. North, “The value of

information visualization,” in Information Visualization - Human-

Centered Issues and Perspectives, ser. Lecture Notes in Computer

Science, A. Kerren, J. T. Stasko, J.-D. Fekete, and C. North, Eds.

Springer, 2008, vol. 4950, pp. 1–18.

[84] J. Stasko, “Value-driven evaluation of visualizations,” in Pro-

ceedings of the Fifth Workshop on Beyond Time and Errors: Novel

Evaluation Methods for Visualization, ser. BELIV ’14. ACM, 2014,

pp. 46–53.

[85] N. Elmqvist, A. V. Moere, H.-C. Jetter, D. Cernea, H. Reiterer, and

T. Jankun-Kelly, “Fluid interaction for information visualization,”

Information Visualization, vol. 10, no. 4, pp. 327–340, 2011.

[86] Y. Wang, A. Segal, R. Klatzky, D. F. Keefe, P. Isenberg, J. Hurti-

enne, E. Hornecker, T. Dwyer, and S. Barrass, “An emotional

response to the value of visualization,” IEEE Computer Graphics

and Applications, vol. 39, no. 5, pp. 8–17, 2019.

[87] Z. Pousman, J. Stasko, and M. Mateas, “Casual information visu-

alization: Depictions of data in everyday life,” IEEE Transactions

on Visualization and Computer Graphics, vol. 13, no. 6, pp. 1145–

1152, 2007.

[88] M. Danziger, “Information visualization for the people,” Mit.edu,

2008.

[89] D. Sprague and M. Tory, “Exploring how and why people use

visualizations in casual contexts: Modeling user goals and reg-

ulated motivations,” Information Visualization, vol. 11, no. 2, pp.

106–123, 2012.

[90] X. Lan, Y. Shi, Y. Zhang, and N. Cao, “Smile or scowl? looking at

infographic design through the affective lens,” IEEE Transactions

on Visualization and Computer Graphics, vol. 27, no. 6, pp. 2796–

2807, 2021.

[91] E. Lee-Robbins and E. Adar, “Affective learning objectives for

communicative visualizations,” IEEE Transactions on Visualization

and Computer Graphics, vol. 29, no. 1, pp. 1–11, 2023.

[92] B. Saket, A. Endert, and J. Stasko, “Beyond usability and per-

formance: A review of user experience-focused evaluations in

visualization,” in Proceedings of the Sixth Workshop on Beyond Time

and Errors on Novel Evaluation Methods for Visualization, ser. BELIV

’16. ACM, 2016, pp. 133–142.

[93] A. Cairo, The Functional Art: An introduction to information graphics

and visualization, ser. Voices That Matter. Pearson Education,

2012.

[94] S. Cohen, “Using visualizations to tell stories,” in The Data

Journalism Handbook. O’Reilly Media, 2011.

[95] S. C. Lewis and O. Westlund, “Big data and journalism,” Digital

Journalism, vol. 3, no. 3, pp. 447–466, 2015.

[96] A. Thudt, J. Walny, T. Gschwandtner, J. Dykes, and J. Stasko,

“Exploration and explanation in data-driven storytelling,” in

Data-Driven Storytelling, N. H. Riche, C. Hurter, N. Diakopoulos,

and S. Carpendale, Eds., 2018, pp. 59–83.

[97] J. Stasko, C. G

org, and Z. Liu, “Jigsaw: Supporting investigative

analysis through interactive visualization,” Information Visualiza-

tion, vol. 7, no. 2, pp. 118–132, 2008.

[98] M. Brehmer, S. Ingram, J. Stray, and T. Munzner, “Overview: The

design, adoption, and analysis of a visual document mining tool

for investigative journalists,” IEEE Transactions on Visualization

and Computer Graphics, vol. 20, no. 12, pp. 2271–2280, 2014.

[99] F. B. Viegas, M. Wattenberg, F. van Ham, J. Kriss, and M. McK-

eon, “ManyEyes: a site for visualization at internet scale,” IEEE

Transactions on Visualization and Computer Graphics, vol. 13, no. 6,

pp. 1121–1128, 2007.

[100] C. Stolte, D. Tang, and P. Hanrahan, “Polaris: a system for

query, analysis, and visualization of multidimensional relational

databases,” IEEE Transactions on Visualization and Computer Graph-

ics, vol. 8, no. 1, pp. 52–65, 2002.

[101] B. Houston, “The history of data journalism: A historical take on

every critical breakthrough from the 1950s until today.”

[102] F. Chevalier, M. Tory, B. Lee, J. v. Wijk, G. Santucci, M. D

ork,

and J. Hullman, “From analysis to communication,” in Data-

Driven Storytelling, N. H. Riche, C. Hurter, N. Diakopoulos, and

S. Carpendale, Eds., 2018, pp. 151–183.

[103] A. Satyanarayan, R. Russell, J. Hoffswell, and J. Heer, “Re-

active Vega: A streaming dataﬂow architecture for declarative

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 18

interactive visualization,” IEEE Transactions on Visualization and

Computer Graphics, vol. 22, no. 1, pp. 659–668, 2016.

[104] B. Lee, R. H. Kazi, and G. Smith, “Sketchstory: Telling more

engaging stories with data through freeform sketching,” IEEE

Transactions on Visualization and Computer Graphics, vol. 19, no. 12,

pp. 2416–2425, 2013.

[105] A. Satyanarayan and J. Heer, “Lyra: An interactive visualization

design environment,” Computer Graphics Forum, vol. 33, no. 3, pp.

351–360, 2014.

[106] Z. Liu, J. Thompson, A. Wilson, M. Dontcheva, J. Delorey,

S. Grigg, B. Kerr, and J. Stasko, “Data Illustrator: Augmenting

vector design tools with lazy data binding for expressive visu-

alization authoring,” in Proceedings of the 2018 CHI Conference on

Human Factors in Computing Systems, ser. CHI ’18. ACM, 2018,

pp. 1–13.

[107] D. Ren, B. Lee, and M. Brehmer, “Charticulator: Interactive

construction of bespoke chart layouts,” IEEE Transactions on

Visualization and Computer Graphics, vol. 25, no. 1, pp. 789–799,

2019.

[108] J. Zong, D. Barnwal, R. Neogy, and A. Satyanarayan, “Lyra

2: Designing interactive visualizations by demonstration,” IEEE

Transactions on Visualization and Computer Graphics, vol. 27, no. 2,

pp. 304–314, 2021.

[109] J. R. Thompson, Z. Liu, and J. Stasko, “Data animator: Authoring

expressive animated data graphics,” in Proceedings of the 2021 CHI

Conference on Human Factors in Computing Systems, ser. CHI ’21.

ACM, 2021.

[110] M. Conlen, M. Vo, A. Tan, and J. Heer, “Idyll studio: A struc-

tured editor for authoring interactive & data-driven articles,” in

The 34th Annual ACM Symposium on User Interface Software and

Technology, ser. UIST ’21. ACM, 2021, pp. 1–12.

[111] Z. Chen and H. Xia, “CrossData: Leveraging text-data connec-

tions for authoring data documents,” in Proceedings of the 2022

CHI Conference on Human Factors in Computing Systems, ser. CHI

’22. ACM, 2022.

[112] Y. Cao, J. L. E, Z. Chen, and H. Xia, “DataParticles: Block-based

and language-oriented authoring of animated unit visualiza-

tions,” in Proceedings of the 2023 CHI Conference on Human Factors

in Computing Systems, ser. CHI ’23. ACM, 2023.

[113] A. Satyanarayan and J. Heer, “Authoring narrative visualizations

with ellipsis,” Computer Graphics Forum, vol. 33, no. 3, pp. 361–

370, 2014.

[114] A. Satyanarayan, B. Lee, D. Ren, J. Heer, J. Stasko, J. Thompson,

M. Brehmer, and Z. Liu, “Critical reﬂections on visualization au-

thoring systems,” IEEE Transactions on Visualization and Computer

Graphics, vol. 26, no. 1, pp. 461–471, 2019.

[115] M. Conlen and J. Heer, “Idyll: A markup language for authoring

and publishing interactive articles on the web,” in Proceedings

of the 31st Annual ACM Symposium on User Interface Software and

Technology, ser. UIST ’18. ACM, 2018, pp. 977–989.

[116] A. Satyanarayan, D. Moritz, K. Wongsuphasawat, and J. Heer,

“Vega-Lite: A grammar of interactive graphics,” IEEE Transactions

on Visualization and Computer Graphics, vol. 23, no. 1, pp. 341–350,

2017.

[117] Y. Fu and J. Stasko, “Supporting data-driven basketball journal-

ism through interactive visualization,” in Proceedings of the 2022

CHI Conference on Human Factors in Computing Systems, ser. CHI

’22. ACM, 2022.

[118] S. Gratzl, A. Lex, N. Gehlenborg, N. Cosgrove, and M. Streit,

“From visual exploration to storytelling and back again,” Com-

puter Graphics Forum, vol. 35, no. 3, pp. 491–500, 2016.

[119] S. Chen, J. Li, G. Andrienko, N. Andrienko, Y. Wang, P. H.

Nguyen, and C. Turkay, “Supporting story synthesis: Bridging

the gap between visual analytics and storytelling,” IEEE Trans-

actions on Visualization and Computer Graphics, vol. 26, no. 7, pp.

2499–2516, 2020.

[120] E. L. Hutchins, J. D. Hollan, and D. A. Norman, “Direct manip-

ulation interfaces,” Human-Computer Interaction, vol. 1, no. 4, pp.

311–338, 1985.

[121] J. Hullman, N. Diakopoulos, E. Momeni, and E. Adar, “Content,

context, and critique: Commenting on a data visualization blog,”

in Proceedings of the 18th ACM Conference on Computer Supported

Cooperative Work amp; Social Computing, ser. CSCW ’15. ACM,

2015, pp. 1170–1175.

[122] N. J. Stroud, E. Van Duyn, and C. Peacock, “News commenters

and news comment readers,” Engaging News Project, pp. 1–21,

2016.

[123] P. J. Boczkowski and E. Mitchelstein, “How users take advantage

of different forms of interactivity on online news sites: Clicking,

e-mailing, and commenting,” Human Communication Research,

vol. 38, no. 1, pp. 1–22, 2012.

[124] E. Dong, H. Du, and L. Gardner, “An interactive web-based

dashboard to track covid-19 in real time.” The Lancet Infectious

Diseases, 2020, vol. 20, no. 5, pp. 533–534.

[125] A. Sarikaya, M. Correll, L. Bartram, M. Tory, and D. Fisher,

“What do we talk about when we talk about dashboards?” IEEE

Transactions on Visualization and Computer Graphics, vol. 25, no. 1,

pp. 682–692, 2019.

[126] M. L. Young, A. Hermida, and J. Fulda, “What makes for great

data journalism?” Journalism Practice, vol. 12, no. 1, pp. 115–135,

2018.

[127] B. Bach, E. Freeman, A. Abdul-Rahman, C. Turkay, S. Khan,

Y. Fan, and M. Chen, “Dashboard design patterns,” IEEE Trans-

actions on Visualization and Computer Graphics, vol. 29, no. 1, pp.

342–352, 2023.

[128] S. Carpendale, “Evaluating information visualizations,” in In-

formation Visualization - Human-Centered Issues and Perspectives,

A. Kerren, J. T. Stasko, J.-D. Fekete, and C. North, Eds. Springer

Berlin Heidelberg, 2008, pp. 19–45.

[129] B. Shneiderman and C. Plaisant, “Strategies for evaluating in-

formation visualization tools: Multi-dimensional in-depth long-

term case studies,” in Proceedings of the 2006 AVI Workshop on

BEyond Time and Errors: Novel Evaluation Methods for Information

Visualization, ser. BELIV ’06. ACM, 2006, pp. 1–7.

[130] M. Sedlmair, M. Meyer, and T. Munzner, “Design study method-

ology: Reﬂections from the trenches and the stacks,” IEEE Trans-

actions on Visualization and Computer Graphics, vol. 18, no. 12, pp.

2431–2440, 2012.

[131] M. Meyer and J. Dykes, “Criteria for rigor in visualization design

study,” IEEE Transactions on Visualization and Computer Graphics,

vol. 26, no. 1, pp. 87–97, 2020.

[132] J. Zimmerman and J. Forlizzi, Research Through Design in HCI,

2014, pp. 167–189.

[133] J. Zimmerman, J. Forlizzi, and S. Evenson, “Research through

design as a method for interaction design research in hci,” in Pro-

ceedings of the SIGCHI Conference on Human Factors in Computing

Systems, ser. CHI ’07. ACM, 2007, pp. 493–502.

[134] D. Domingo, “Interactivity in the daily routines of online

newsrooms: dealing with an uncomfortable myth,” Journal of

Computer-Mediated Communication, vol. 13, no. 3, pp. 680–704,

2008.

[135] T. Munzner, “Process and pitfalls in writing information visu-

alization research papers,” in Information Visualization: Human-

Centered Issues and Perspectives, A. Kerren, J. T. Stasko, J.-D. Fekete,

and C. North, Eds. Springer Berlin Heidelberg, 2008, pp. 134–

153.

[136] F. Amini, N. H. Riche, B. Lee, A. Monroy-Hernandez, and P. Irani,

“Authoring data-driven videos with dataclips,” IEEE Transactions

on Visualization and Computer Graphics, vol. 23, no. 1, pp. 501–510,

2017.

[137] Z. Chen, S. Ye, X. Chu, H. Xia, H. Zhang, H. Qu, and Y. Wu,

“Augmenting sports videos with viscommentator,” IEEE Trans-

actions on Visualization and Computer Graphics, vol. 28, no. 1, pp.

824–834, 2022.

[138] M. Shin, J. Kim, Y. Han, L. Xie, M. Whitelaw, B. C. Kwon, S. Ko,

and N. Elmqvist, “Roslingiﬁer: Semi-automated storytelling for

animated scatterplots,” IEEE Transactions on Visualization and

Computer Graphics, vol. 29, no. 6, pp. 2980–2995, 2023.

[139] Z. Chen, Q. Yang, X. Xie, J. Beyer, H. Xia, Y. Wu, and H. Pﬁster,

“Sporthesia: Augmenting sports videos using natural language,”

IEEE Transactions on Visualization and Computer Graphics, vol. 29,

no. 1, pp. 918–928, 2023.

[140] T. Tang, J. Tang, J. Hong, L. Yu, P. Ren, and Y. Wu, “Design

guidelines for augmenting short-form videos using animated

data visualizations,” Journal of Visualization, vol. 23, no. 4, pp.

707–720, 2020.

[141] T. Lin, Z. Chen, Y. Yang, D. Chiappalupi, J. Beyer, and H. Pﬁster,

“The quest for : Embedded visualization for augmenting basket-

ball game viewing experiences,” IEEE Transactions on Visualization

and Computer Graphics, vol. 29, no. 1, pp. 962–971, 2023.

[142] M. Wattenberg, “Baby names, visualization, and social data anal-

ysis,” in IEEE Symposium on Information Visualization, ser. InfoVis

’05, 2005, pp. 1–7.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 19

[143] B. Lee, E. K. Choe, P. Isenberg, K. Marriott, and J. Stasko, “Reach-

ing broader audiences with data visualization,” IEEE Computer

Graphics and Applications, vol. 40, no. 2, pp. 82–90, 2020.

[144] B. Lee, A. Srinivasan, J. Stasko, M. Tory, and V. Setlur, “Mul-

timodal interaction for data visualization,” in Proceedings of the

2018 International Conference on Advanced Visual Interfaces, ser. AVI

’18. ACM, 2018.

[145] B. Lee, A. Srinivasan, P. Isenberg, and J. Stasko, “Post-WIMP

interaction for information visualization,” Foundations and Trends

in Human-Computer Interaction, vol. 14, no. 1, pp. 1–95, 2021.

[146] S. Reeves, S. Benford, C. O’Malley, and M. Fraser, “Designing the

spectator experience,” in Proceedings of the SIGCHI Conference on

Human Factors in Computing Systems, ser. CHI ’05. ACM, 2005,

pp. 741–750.

[147] Q. Zhi, S. Lin, P. Talkad Sukumar, and R. Metoyer, “Gameviews:

Understanding and supporting data-driven sports storytelling,”

in Proceedings of the 2019 CHI Conference on Human Factors in

Computing Systems, ser. CHI ’19. ACM, 2019, pp. 1–13.

[148] C. G

org, Z. Liu, J. Kihm, J. Choo, H. Park, and J. Stasko, “Com-

bining computational analyses and interactive visualization for

document exploration and sensemaking in Jigsaw,” IEEE Trans-

actions on Visualization and Computer Graphics, vol. 19, no. 10, pp.

1646–1663, 2013.

[149] C. G

org, Z. Liu, and J. Stasko, “Reﬂections on the evolution of the

Jigsaw visual analytics system,” Information Visualization, vol. 13,

no. 4, pp. 336–345, 2014.

[150] Y.-a. Kang, C. G

org, and J. Stasko, “How can visual analytics

assist investigative analysis? Design implications from an evalu-

ation,” IEEE Transactions on Visualization and Computer Graphics,

vol. 17, no. 5, pp. 570–583, 2011.

[151] T. Erickson and W. A. Kellogg, “Social translucence,” ACM

Transactions on Computer-Human Interaction, vol. 7, no. 1, pp. 59–

83, 2000.

[152] M. Eslami, A. Rickman, K. Vaccaro, A. Aleyasen, A. Vuong,

K. Karahalios, K. Hamilton, and C. Sandvig, “”I always assumed

that I wasn’t really that close to [her]”: Reasoning about invisible

algorithms in news feeds,” in Proceedings of the 33rd Annual ACM

Conference on Human Factors in Computing Systems, ser. CHI ’15.

ACM, 2015, pp. 153–162.

[153] J. Choo and S. Liu, “Visual analytics for explainable deep learn-

ing,” IEEE Computer Graphics and Applications, vol. 38, no. 4, pp.

84–92, 2018.

[154] W. Samek, T. Wiegand, and K.-R. M

uller, “Explainable artiﬁcial

intelligence: Understanding, visualizing and interpreting deep

learning models,” arXiv preprint arXiv:1708.08296, 2017.

[155] F. Hohman, M. Kahng, R. Pienta, and D. H. Chau, “Visual

analytics in deep learning: An interrogative survey for the next

frontiers,” IEEE Transactions on Visualization and Computer Graph-

ics, vol. 25, no. 8, pp. 2674–2693, 2019.

[156] Z. J. Wang, R. Turko, O. Shaikh, H. Park, N. Das, F. Hohman,

M. Kahng, and D. H. Polo Chau, “Cnn explainer: Learning

convolutional neural networks with interactive visualization,”

IEEE Transactions on Visualization and Computer Graphics, vol. 27,

no. 2, pp. 1396–1406, 2021.

[157] J. Burrell, “How the machine ‘thinks’: Understanding opacity in

machine learning algorithms,” Big Data & Society, vol. 3, no. 1,

2016.

[158] M. T. Ribeiro, S. Singh, and C. Guestrin, “”Why should I trust

you?”: Explaining the predictions of any classiﬁer,” in Proceedings

of the 22nd ACM SIGKDD International Conference on Knowledge

Discovery and Data Mining, ser. KDD ’16. ACM, 2016, pp. 1135–

1144.

[159] D. Lupton, “Feeling your data: Touch and making sense of

personal digital data,” New Media & Society, vol. 19, no. 10, pp.

1599–1614, 2017.

[160] G. Lupi, S. Posavec, and M. Popova, Dear Data. Princeton

Architectural Press, 2016.

[161] J. Kastrenakes and J. Peters, “Webby awards

2020: the complete winners list,” https:

//www.theverge.com/2020/5/20/21263445/

2020-webby-awards-winners-lil-nas-x-nasa-jon-krasinski,

The Verge, May 2020.

[162] “Story portrait,” https://help.nytimes.com/hc/en-us/articles/

6921803704468, The New York Times, 2020.

[163] S. Munson, S. Lee, and P. Resnick, “Encouraging reading of

diverse political viewpoints with a browser widget,” Proceedings

of the International AAAI Conference on Web and Social Media, vol. 7,

no. 1, pp. 419–428, 2021.

[164] G. R. Hayes, “The relationship of action research to human-

computer interaction,” ACM Transactions on Computer-Human

Interaction, vol. 18, no. 3, pp. 1–20, 2011.

[165] G. Hayes, Knowing by Doing: Action Research as an Approach to

HCI. Springer New York, 2014, pp. 49–68.

[166] VisualizeNews, “India goes to polls,” https://india.visualize.

news/, Indian General Elections 2019, 2019.

[167] N. Nigro, “Hamilton 2.0 dashboard,” https://

securingdemocracy.gmfus.org/hamilton-dashboard/, Alliance

For Securing Democracy, 2022.

[168] L. Nicoletti and S. Sarva, “When women make headlines: A

visual essay about the (mis)representation of women in the

news,” https://pudding.cool/2022/02/women-in-headlines/,

The Pudding, 2021.

[169] N. McCurdy, J. Dykes, and M. Meyer, “Action design research

and visualization design,” in Proceedings of the Sixth Workshop on

Beyond Time and Errors on Novel Evaluation Methods for Visualiza-

tion, ser. BELIV ’16. ACM, 2016, pp. 10–18.

[170] J. Zhou, Y. Zhang, Q. Luo, A. G. Parker, and M. De Choudhury,

“Synthetic lies: Understanding ai-generated misinformation and

evaluating algorithmic and human solutions,” in Proceedings of

the 2023 CHI Conference on Human Factors in Computing Systems,

ser. CHI ’23. ACM, 2023.

[171] K. Shu, D. Mahudeswaran, and H. Liu, “Fakenewstracker: a tool

for fake news collection, detection, and visualization,” Compu-

tational and Mathematical Organization Theory, vol. 25, no. 1, pp.

60–71, 2019.

[172] A. Karduni, I. Cho, R. Wesslen, S. Santhanam, S. Volkova, D. L.

Arendt, S. Shaikh, and W. Dou, “Vulnerable to misinformation?

veriﬁ!” in Proceedings of the 24th International Conference on Intelli-

gent User Interfaces, ser. IUI ’19. ACM, 2019, pp. 312–323.

[173] A. Karduni, R. Wesslen, S. Santhanam, I. Cho, S. Volkova,

D. Arendt, S. Shaikh, and W. Dou, “Can you veriﬁ this? studying

uncertainty and decision-making about misinformation using

visual analytics,” Proceedings of the International AAAI Conference

on Web and Social Media, vol. 12, no. 1, Jun. 2018.

[174] S. Lee, S. Afroz, H. Park, Z. J. Wang, O. Shaikh, V. Sehgal,

A. Peshin, and D. H. Chau, “Misvis: Explaining web misinfor-

mation connections via visual summary,” in Extended Abstracts of

the 2022 CHI Conference on Human Factors in Computing Systems,

ser. CHI EA ’22. ACM, 2022.

[175] D. Lazer, M. Baum, N. Grinberg, L. Friedland, K. Joseph,

W. Hobbs, and C. Mattsson, “Combating fake news: An agenda

for research and action,” 2017.

[176] H. Rashkin, E. Choi, J. Y. Jang, S. Volkova, and Y. Choi, “Truth of

varying shades: Analyzing language in fake news and political

fact-checking,” in Proceedings of the 2017 Conference on Empirical

Methods in Natural Language Processing. Association for Compu-

tational Linguistics, Sep. 2017, pp. 2931–2937.

[177] E. Wall, A. Narechania, A. Coscia, J. Paden, and A. Endert, “Left,

right, and gender: Exploring interaction traces to mitigate human

biases,” IEEE Transactions on Visualization and Computer Graphics,

vol. 28, no. 1, pp. 966–975, 2022.

[178] A. Narechania, A. Coscia, E. Wall, and A. Endert, “Lumos:

Increasing awareness of analytic behavior during visual data

analysis,” IEEE Transactions on Visualization and Computer Graph-

ics, vol. 28, no. 1, pp. 1009–1018, 2022.

[179] A. D. Holan, “The principles of the Truth-O-Meter:

PolitiFact’s methodology for independent fact-checking,”

https://www.politifact.com/article/2018/feb/12/

principles-truth-o-meter-politifacts-methodology-i/, Politifact,

2020.

[180] C. R. Sunstein, S. C. Lazzaro, and T. Sharot, “How people update

beliefs about climate change: Good news and bad news,” SSRN

Electronic Journal, 2016.

[181] A. V. Pandey, K. Rall, M. L. Satterthwaite, O. Nov, and E. Bertini,

“How deceptive are deceptive visualizations? an empirical anal-

ysis of common distortion techniques,” in Proceedings of the 33rd

Annual ACM Conference on Human Factors in Computing Systems,

ser. CHI ’15. ACM, 2015, pp. 1469–1478.

[182] A. Fan, Y. Ma, M. Mancenido, and R. Maciejewski, “Annotating

line charts for addressing deception,” in Proceedings of the 2022

CHI Conference on Human Factors in Computing Systems, ser. CHI

’22. ACM, 2022, pp. 1–12.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.

JOURNAL OF L

X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 20

[183] A. McNutt, G. Kindlmann, and M. Correll, “Surfacing visualiza-

tion mirages,” in Proceedings of the 2020 CHI Conference on Human

Factors in Computing Systems, ser. CHI ’20. ACM, 2020, pp. 1–16.

[184] A. K. Hopkins, M. Correll, and A. Satyanarayan, “VisuaLint:

Sketchy in situ annotations of chart construction errors,” Com-

puter Graphics Forum, vol. 39, no. 3, pp. 219–228, 2020.

[185] The White House, “We just learned that President Biden’s

ﬁrst year in ofﬁce was the strongest year for economic

growth since 1984,” https://twitter.com/WhiteHouse/status/

1486709480351952901, 2022.

[186] B. E. Rogowitz, L. A. Treinish, and S. Bryson, “How not to lie

with visualization,” Computers in Physics, vol. 10, no. 3, p. 268,

1996.

[187] E. R. Tufte, The Visual Display of Quantitative Information, 2nd ed.

Graphics Press, 2001.

[188] M. A. Borkin, Z. Bylinskii, N. W. Kim, C. M. Bainbridge, C. S.

Yeh, D. Borkin, H. Pﬁster, and A. Oliva, “Beyond memorability:

Visualization recognition and recall,” IEEE Transactions on Visual-

ization and Computer Graphics, vol. 22, no. 1, pp. 519–528, 2016.

[189] H.-K. Kong, Z. Liu, and K. Karahalios, “Frames and slants in

titles of visualizations on controversial topics,” in Proceedings of

the 2018 CHI Conference on Human Factors in Computing Systems,

ser. CHI ’18. ACM, 2018, pp. 1–12.

[190] ——, “Trust and recall of information across varying degrees of

title-visualization misalignment,” in Proceedings of the 2019 CHI

Conference on Human Factors in Computing Systems, ser. CHI ’19.

ACM, 2019, pp. 1–13.

[191] P.-M. Law, A. Endert, and J. Stasko, “Characterizing automated

data insights,” in 2020 IEEE Visualization Conference (VIS), 2020,

pp. 171–175.

[192] C. Oh, J. Choi, S. Lee, S. Park, D. Kim, J. Song, D. Kim, J. Lee,

and B. Suh, “Understanding user perception of automated news

generation system,” in Proceedings of the 2020 CHI Conference on

Human Factors in Computing Systems, ser. CHI ’20. ACM, 2020,

pp. 1–13.

[193] Y. Wang, Z. Sun, H. Zhang, W. Cui, K. Xu, X. Ma, and D. Zhang,

“DataShot: Automatic generation of fact sheets from tabular

data,” IEEE Transactions on Visualization and Computer Graphics,

vol. 26, no. 1, pp. 895–905, 2020.

[194] C. Bryan, K.-L. Ma, and J. Woodring, “Temporal summary

images: An approach to narrative visualization via interactive

annotation generation and placement,” IEEE Transactions on Visu-

alization and Computer Graphics, vol. 23, no. 1, pp. 511–520, 2017.

[195] D. Shi, X. Xu, F. Sun, Y. Shi, and N. Cao, “Calliope: Automatic vi-

sual data story generation from a spreadsheet,” IEEE Transactions

on Visualization and Computer Graphics, vol. 27, no. 2, pp. 453–463,

2021.

[196] A. Srinivasan, S. M. Drucker, A. Endert, and J. Stasko, “Augment-

ing visualizations with interactive data facts to facilitate interpre-

tation and communication,” IEEE Transactions on Visualization and

Computer Graphics, vol. 25, no. 1, pp. 672–681, 2019.

[197] I. Bogost, S. Ferrari, and B. Schweizer, Newsgames: Journalism at

Play, ser. The MIT Press. MIT Press, 2012.

[198] N. Diakopoulos, F. Kivran-Swaine, and M. Naaman, Playable data:

characterizing the design space of game-y infographics, 2011.

[199] P. Isenberg, B. Lee, H. Qu, and M. Cordeil, “Immersive visual

data stories,” in Immersive Analytics, K. Marriott, F. Schreiber,

T. Dwyer, K. Klein, N. H. Riche, T. Itoh, W. Stuerzlinger, and

B. H. Thomas, Eds. Cham: Springer International Publishing,

2018, pp. 165–184.

[200] M. Broussard, “Archiving data journalism.” O’Reilly Media,

2012.

[201] K. Boss and M. Broussard, “Challenges of archiving and preserv-

ing born-digital news applications,” IFLA Journal, vol. 43, no. 2,

pp. 150–157, 2017.

[202] R. Kosara, “The bits are rotting in the state of

data journalism,” https://eagereyes.org/blog/2016/

the-bits-are-rotting-in-the-state-of-data-journalism, eagereyes,

Jul 2016.

[203] M. Broussard, “Preserving news apps present huge challenges,”

Newspaper Research Journal, vol. 36, no. 3, pp. 299–313, 2015.

[204] M. Broussard and K. Boss, “Saving data journalism,” Digital

Journalism, vol. 6, no. 9, pp. 1206–1221, 2018.

[205] L. Ahva and S. Steensen, “Journalism theory,” in The Handbook of

Journalism Studies, 2nd ed., K. Wahl-Jorgensen and T. Hanitzsch,

Eds. Routledge, 2019, pp. 38–54.

[206] E. M. Peck, S. E. Ayuso, and O. El-Etr, “Data is personal: Attitudes

and perceptions of data visualization in rural pennsylvania,”

in Proceedings of the 2019 CHI Conference on Human Factors in

Computing Systems, ser. CHI ’19. ACM, 2019, pp. 1–12.

[207] B. Alper, N. H. Riche, F. Chevalier, J. Boy, and M. Sezgin, “Visu-

alization literacy at elementary school,” in Proceedings of the 2017

CHI Conference on Human Factors in Computing Systems, ser. CHI

’17. ACM, 2017, pp. 5485–5497.

[208] M. Correll, “Position paper: Are we making progress in visualiza-

tion research?” in 2022 IEEE Evaluation and Beyond - Methodological

Approaches for Visualization (BELIV), 2022, pp. 1–10.

Yu Fu is a Ph.D. student at Georgia Tech’s

School of Interactive Computing and a member

of the Information Interfaces Group. His research

lies at the intersection of information visualiza-

tion and journalism, with a focus on leveraging

visualization to help journalists and their audi-

ences cope with emerging challenges. He re-

ceived his B.S. and M.S. in Electrical Engineer-

ing. He also worked as a journalist.

John Stasko is a Regents Professor in the

School of Interactive Computing and the Director

of the Information Interfaces Research Group

at the Georgia Institute of Technology. His re-

search is in the areas of information visualiza-

tion and visual analytics, approaching each from

a human-computer interaction perspective. He

was named an IEEE Fellow in 2014 and an ACM

Fellow in 2022.

This article has been accepted for publication in IEEE Transactions on Visualization and Computer Graphics. This is the author's version which has not been fully edited and

content may change prior to final publication. Citation information: DOI 10.1109/TVCG.2023.3287585

Authorized licensed use limited to: Hong Kong University of Science and Technology. Downloaded on July 19,2023 at 07:43:48 UTC from IEEE Xplore. Restrictions apply.